Citation
Some new chemistry of pentacoordinate organofluorosilanes

Material Information

Title:
Some new chemistry of pentacoordinate organofluorosilanes
Creator:
Danahey, Stephen Edward
Publication Date:
Language:
English
Physical Description:
xi, 148 leaves : illustrations ; 29 cm

Subjects

Subjects / Keywords:
Organofluorosilane compounds ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 138-145).
General Note:
Submitted in partial fulfillment of the requirements for the degree of Master of Science, Department of Chemistry
Statement of Responsibility:
by Stephen Edward Danahey.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
17995267 ( OCLC )
ocm17995267
Classification:
LD1190.L46 1986m .D36 ( lcc )

Full Text
SOME NEW CHEMISTRY OF PENTACOORDINATE
ORGANOFLUOROSILANES
by
Stephen Edward Danahey
B.A., University of Colorado-Denver, 1980
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
1986
w \
r -S


This thesis for the Master of Science degree by
Stephen Edward Danahey
has been approved for the
Department of
Chemistry
Date______71^3 (8b


iii
Danahey, Stephen Edward (M.S., Chemistry)
Some New Chemistry of Pentacoordinate Organofluorosilanes
Thesis directed by Professor Robert Damrauer
Fluoride ion solubilized by 18-crown-6 has been
reacted with a variety of chloromethyl-substituted silanes,
which have been shown to undergo alkyl and aryl
rearrangement. The competitive migration of various groups
demonstrated that the ability of the migrating group to sup-
port negative charge controls the rearrangement.
The use of the silacrown ethers 1,1-dimethylsila-11-
crown-J*, 1,1-dimethylsila-1 4-crown-5 and 1,1-dimethylsila-17~
crown-6 as phase-transfer catalysts for fluoride ion has been
examined. The silacrowns have proven unsuitable as a sub-
stitute for l8-crown-6.
The l8-crown-6 potassium salts of
trifluorodiphenylsiliconate, trifluoromethylphenylsiliconate,
tetrafluorophenylsiliconate, difluorotriphenylsiliconate,
bis(m-fluorophenyl)phenyldifluorosiliconate, tris(m-
fluorophenyl)difluorosiliconate, trifluorodixylylsiliconate
and tetrafluoroxylylsiliconate have been prepared. Their
dynamic NMR behavior has been examined in detail. Addition-
ally, the trifluorodixylylsiliconate decomposes by an unknown
pathway to give the tetrafluoroxylylsiliconate.


Examine the contents, not the bottle.
From the Talmud
If I had it to do over, I would do it
better, maybe even right.
William Faulkner
When I began to write myself, my
brother encouraged me and he gave me
certain rules for writing. He said:
when you write tell a story, and
don't try to explain the story. You
tell [the reader] the story, and the
explanations and interpretations
he will make himself, or critics will
do it for him.
Isaac Bashevis Singer


ACKNOWLEDGEMENTS
Louis Aran; Bonnie K. O'Connell. Vern Yost and Roger
Simon. Jim Crofter and Mike Milash for much neededand
frequentassistancei My mentor, Bob Damrauer. Although a modern
man, he has many of the fine qualities of the people in the
pictures of his office wall.


CONTENTS
CHAPTER
I. INTRODUCTION.......................................... 1
II. THE REACTION OF SOLUBILIZED FLUORIDE ION WITH
CHLOROMETHYL-SUBSTITUTED SILANES..................... 4
2.1 Introduction and Review.......................... ^
2.2 Experimental Results............................. 9
2.2.1 Reaction of Me3SiCH2Cl with KF............ 9
2.2.2 Reaction of Me3SiCH2Cl with KI............ 10
2.2.3 Comparison of the Reactivity of Me3SiCH2Cl
with CsF, with KF and with NaF.............. 10
2.2.4 Reaction with no l8-Crown-6................. 11
2.2.5 Reaction with no Solvent.................... 11
2.2.6 Water Content of Oven-dried (120 C) KF... 12
2.2.7 Reaction of Me3SiCH2Cl with n-BuHN+F ....... 13
2.2.8 Reaction of M|3SiCH2Cl with KF and
MeN(n-octyl)3 Cl .......................... 13
2.2.9 Reaction of Me3SiCH2Cl with KF
in Nonaromatic Solvents...................... 14
2.2.10 Reaction using Stoichiometric l8-Crown-6.. 14
2.2.11 Reaction of Me2PhSiCH2Cl with KF............. 15
2.2.12 Reaction of Me2PhSiCH2Cl with CsF........... 16
2.2.13 Reaction of Me2(CH2=CH)SiCH2Cl with KF____ 16
2.2.14 Preparation of Me2(F)SiCH2Cl................. 17
2.2.15 Reaction of Me2(F)SiCH2Cl with KF............ 18


vii
2.2.16 Reaction of Me3SiCH(Cl)CH3 with KF........ 19
2.3 Discussion........................................ 20
2.3.1 Reaction of Me3SiCH2Cl with KF and KI..... 20
2.3.2 Various Reagents and Reaction Conditions.. 22
2.3.3 Reaction of Me2RSiCH2Cl
(R = Ph and CH2=CH)....................... 23
2.3.4 More Definitive Studies................... 23
2.3.5 Reaction of Me2(F)SiCH2Cl.................... 24
2.3*6 No Reaction with Me3SiCH(Cl)CH3................ 25
III. AN INVESTIGATION OF
SILACROWN ETHERS.................................. 26
3.1 Introduction and Review......................... 26
3.2 Experimental Results............................ 29
3.2.1 Reaction of Me3SiCH2Cl with KF
and Silacrown-4.............................. 29
3.2.2 Reaction of Me3SiCH2Cl with KF
and Silacr own-5............................. 29
3.2.3 Reaction of Me3SiCH2Cl with KF
and Silacr own-6............................. 30
3.2.4 Reaction of Me3SiCH2Cl with CsF
and Silacrown-4.............................. 31
3.2.5 Reaction of Me3SiCH2Cl with CsF
and Silacrown-5.............................. 31
3.2.6 Reaction of Me3SiCH2Cl with CsF
and Silacr own-6............................. 32
3.2.7 Solubility Studies: Silacrowns vs.
18-Crown-6................................. 32
3.2.8 19F NMR of KF and l8-Crown-6................ .33
3.2.9 19F NMR of KF and Silacrown-6
(with Ph2SiF2)............................... 33


viii
3.2.10 19F NMR of KF and Silacrown-5
(with Ph2SiF2)............................ 34
3.2.11 19F NMR of KF and Silacrown-4 (with
Ph2SiF2 and MePhSiF2)..................... 34
3.2.12 19F NMR of Ph2SiF2 and MePhSiF2........... 35
3.3 Discussion................................... 36
3.3.1 Reaction of Me3SiCH2Cl with F and
Silacrown................................. 36
3.3.2 19F NMR................................... 36
IV. THE REACTION OF FLUORIDE ION WITH 0RGAN0FLU0R0-
SILANES TO FORM PENTACOORDINATE SILICON ANIONS.... 38
4.1 Introduction and Review...................... 38
4.1.1 Pentacoordinate SiliconEarly Work........ 40
4.1.2 Pentacoordinate SiliconRecent Work......... 49
4.1.3 Propeller Compounds....................... 61
4.2 Experimental Results......................... 66
4.2.1 NMR Data for Ph2SiF2...................... 66
4.2.2 Reaction of Ph2SiF2 with KF
and l8-Crown-6............................ 66
4.2.3 Low Temperature NMR Data for Ph2SiF3 .. 68
4.2.4 NMR Data for MePhSiF2..................... 70
4.2.5 Reaction of MePhSiF2 with KF
and 18-Crown-6........................... 70
4.2.6 Low Temperature NMR Data for MePhSiF3 .... 72
4.2.7 Simulation of the 19F_NMR Spectra of
Ph2SiF3 and MePhSiF3 .................... 74
4.2.8 Preparation of PhSiF3..................... 80
4.2.9 Reaction of PhSiF3 with KF
and l8-Crown-6............................ 81


ix
4.2.10 Low Temperature NMR Spectra of PhSiF* .... 82
4.2.11 NMR Data for Ph3SiF.......................... 82
4.2.12 Reaction of Ph3SiF with KF
and 18-Crown-6.............................. 83
4.2.13 19F NMR Spectrum of a Ph3SiF2 /Ph3SiF
Mixture....................................... 84
4.2.14 Low Temperature NMR Data for Ph3SiF2 ........ 85
4.2.15 NMR Data for (m-FPh) 2PhSiF................... 85
4.2.16 Low Temperature NMR Data
for (m-FPh )2PhSiF............................ 86
4.2.17 Reaction of (m-FPh)2PhSiF with KF
and l8-Crown-6................................ 86
4.2.18 Low Temperature NMR Data
for (m-FPh)2PhSiF2 ........................... 87
4.2.19 NMR Data for (m-FPh)3SiF...................... 88
4.2.20 Low Temperature NMR Data for (m-FPh)3SiF.. 88
4.2.21 Reaction of (m-FPh)3SiF with KF
and l8-Crown-6................................ 89
4.2.22 Low Temperature NMR Data for
(m-FPh)3SiF2 ................................. 90
4.2.23 Preparation of o-Xylyl2SiF2................... 90
4.2.24 Low Temperature NMR Data
for o-Xylyl2SiF2.............................. 92
4.2.25 Reaction of o-Xylyl2SiF2 with KF
and 18-Crown-6................................ 92
4.2.26 Low Temperature NMR Data
for o-Xylyl2SiF3~............................. 94
4.2.27 Crude Estimation of the Activation
Parameters for o-Xylyl2SiF3 .................. 96
4.2.28 Disproportionation of o-Xylyl2SiF3
to o-XylylSiF,, and o-Xylene-d!............... 98


X
4.2.29 NMR Data for o-XylylSiF3.................. 99
4.2.30 Reaction of o-XylylSiF3 with KF
and l8-Crown-6............................ 100
4.2.31 Low Temperature_NMR Data
for o-XylylSiF* ......................... 102
4.2.32 Reaction of SiCl* with KF
and l8-Crown-6............................ 102
4.2.33 Preparation of HPhSiF2.................... 102
4.2.34 Reaction of HPhSiF2 with KF
and l8-Crown-6.............................. 103
4.2.35 Low Temperature NMR Data
for "HPhSiF3~".............................. 105
4.2.36 NMR Data for (Trisyl)SiPh2F =
(Me3Si)3CSiPh2F............................. 105
4.2.37 Reaction of (Trisyl)SiPh2F with KF
and l8-Crown-6........................... 106
4.2.38 NMR Data for 1,1-Difluorosilacyclobutane,
(CH2)3SiF2.................................. 107
4.2.39 Reaction of (CH2)3SiF2 with KF
and l8-Crown-6.............................. 108
4.2.40 Preparation of 1-Fluoro-1-methylsila-
cyclobutane, (CH2)3Si(F)Me.................. 109
4.2.41 Reaction of (CH2)3Si(F)Me with KF
and 1 8-Crown-6............................. 110
4.2.42 Other Reaction Attempts with KF
and l8-Crown-6............................. 111
4.3 Discussion................................... 112
4.3.1 Ph2SiF3 (VII).............................. 116
4.3.2 MePhSiF3" (VIII)............................ 119
4.3.3 Activation Barriers for Ph2SiF3 (VII)
and MePhSiF3 (VIII)......................... 123
4.3.4 PhSiF/ (IX)................................. 124


xi
4.3.5 Ph3SiF2~ (X), (m-FPh)2PhSiF2 (XI),
and (m-FPh)3SiF2 (XII)....................... 126
4.3.6 o-Xylyl2SiF^" (XIII) and
0-XylylSiF (XIV)............................ 127
4.3.7 Chemical Shifts............................. 132
4.3.8 Coupling Constants........................... 136
REFERENCES AND NOTES....................................... 138
APPENDIX..................................................... 146
C


TABLES
TABLE
I. Bond energies......................................... 1
II. NMR data of four-coordinate starting silanes....... 49
III. Spiro siliconates................................. 58
IV. Temperature-dependent behavior of Ph2SiF3 ........... 69
V. Temperature-dependent behavior of MePhSiF3 .......... 73
VI. CBS Analysis for Ph2SiF3 and MePhSiF3 ............... 77
VII. Activation parameters for Ph2SiF3 and MePhSiF3 .. 77
VIII. Temperature-dependent behavior of o-Xylyl3SiF2 ... 95
IX. Estimation of AG* for o-Xylyl2SiF3 .................. 97
X. Stepwise Substitution in Siliconates/Phosphoranes. 113
XI. Triphenyl Siliconates............................... 115
XII. Mono- and Diphenyl Siliconates.................... 115


FIGURES
FIGURE
I. Plot of log k vs. 1/T.............................. 78
II. Plot of log k/T vs. 1/T............................ 79
III. Plot of the absolute value of the 19F chemical
shift vs. the negative of the sum of the
Taft values for the ligands on silicon................... 131*
IV. Plot of the absolute value of the 19F chemical
shift vs. the negative of the sum of the
Taft (Tj values for the ligands on
pentacoordinate phosphorus.......................,.. 135


CHAPTER I
INTRODUCTION
The Si-F bond is unique because it is one of the
strongest known single bonds. For comparison, a quantitative
picture of some silicon and carbon single bond strengths1 is
given in Table I. Because such a strong bond is being
Table I. Bond energies (kcal/mol)
Silicon Bond Bond Energy Carbon Bond Bond Energy
Si-Si 7^ C-Si 88
Si-C 88 c-c 88
Si-H 90 C-H 98
Si-0 128 C-0 91
Si-N 100 C-N 80
Si-F 154 C-F 11-6
Si-Cl 113 C-Cl 80
Si-Br 96 C-Br 64
Si-I 77 C-I 51
made, nucleophilic attack at silicon is particularly easy with
fluoride ion.


2
Silicon holds a unique and special position in the
hierarchy of elements. Its chemistry can be viewed as lying
between that of organic and inorganic chemistry. Typically,
silicon has a coordination number of four and many silicon
compounds are similar to saturated carbon compounds. Silicon
also resembles other elements in that it does not form stable
multiple-bonded molecules, there are relatively few catenated
silicon species and it has the.ability to expand its coordina-
tion sphere. In fact, hypervalent silicon anions are fre-
quently postulated as intermediates in the reactions of or-
ganosilicon compounds with nucleophiles.2 Such an inter-
mediate (most often regarded as a pentacovalent trigonal
bipyramidal silicon) is a distinct entity, which by a series
of pseudorotations3 can rearrange to another (geometrically
distinct) intermediate. This is, of course, quite different
from all forms of the S2 transition state of carbon
N
chemistry.
The goal of this work was to use Si-F bond formation
to learn more about the mechanisms and particularly the inter-
mediates of organosilicon reactions.
Our group has used cation-binding crown ethers, which
serve as solid-to-liquid phase transfer catalysts for potas-
sium fluoride, to study the nucleophilic attack of solubilized
fluoride ion on chloromethylsilanes. That study, summarized
in. Chapter II, initiated our interest in solution-phase pen-


3
tacoordinate silicon anions, which in most circumstances have
lifetimes too short for identification.
We have attempted to rigorously identify silicon
anions in the interaction of potassium fluoride with some
unusual phase transfer catalysts, the silicon analogues of
ring-contracted crown ethers. We also wished to determine
whether these unusual catalysts offer any advantage over more
typical crown ethers. This work is outlined in Chapter III.
In addition, we have discovered a new method for
preparing pentacoordinate compounds from reaction of
fluorosilanes with solubilized fluoride ion. A study of the
structures and reactivities of some of these species is
presented in Chapter IV.


CHAPTER II
THE REACTION OF SOLUBILIZED FLUORIDE ION WITH
CHLOROMETHYL-SUBSTITUTED ORGANOSILANES
2.1 Introduction and Review
Previous work has demonstrated that
chloromethylorganosilanes (R3SiCH2Cl) react with anionic
reagents in a complex manner. Products result from
nucleophilic attack at the carbon center of the chloromethyl
group and at the silicon center. For example, the reaction of
sodium ethoxide with chloromethyldimethylphenylsilane1*
demonstrated three principal reaction pathways. These consist
of substitution at the carbon of the chloromethyl group,
elimination (cleavage), of the chloromethyl group and 1,2
migration or rearrangement of phenyl anion from silicon to
carbon (eq 1). Eaborn has suggested5 that the common
Me2PhSiCH2Cl
substitution
- )
Eto" cleavage
EtOH rearrangement
Me2PhSiCH2OEt
42?
Me2PhSiOEt
16?
Me 2(PhCH 2)Si OEt
Me2Si(OEt.)2
32?
(1)


5
intermediate in all these pathways is a pentacovalent silicon
anion.
We have shown6 that the silicon analog of neopentyl
chloride, chloromethyltrimethylsilane, reacts in aromatic
solvents with potassium or cesium fluoride in the presence of
l8-crown-6 to give exclusively the 1,2 methyl migration or
rearrangement product, dimethylethylfluorosilane (eq 2):
KF or CsF
18-crown-6
(CH3)3SiCH2Cl ----------- (CH3)2(CH3CH2)SiF (2)
Toluene
In addition, chloromethyldimethyl aryl-substituted
silanes undergo competitive alkyl and aryl rearrangement under
the same conditions (eq 3)- We have studied the competitive
+ (CH3)2(ArCH2)SiF
ma j or
primary product
+ (3)
CH3(Ar)(CH3CH2)SiF
minor
primary product
migration of various Ar (and R) groups and determined, using
primary product distributions, that migratory aptitudes of
simple aryl and alkyl groups are controlled by the ability of
(CH3)2ArSiCH2Cl
KF
18-crown-6
Aromatic
Solvent


6
the migrating group to bear the negative charge which develops
in the transition state.6
Related facile rearrangements of chloromethyl-
substituted silanes under electrophilic conditions have been
known for a long time (eq 4).7 These are similar to the
A1C13
(CH3)3SiCH2Cl -------------> (CH3)2(CH3CH2)SiCl (4)
Wagner-Meerwein rearrangement of carbon chemistry and are
strongly dependent upon the build-up of positive charge at the
chloromethyl carbon center. They are not complicated,
however, by the possibilty of competitive migrations.8
Only a few examples of nucleophile induced rearrange-
ments of chloromethyl-substituted silanes are known. These
include the elegant fluoride-induced phenyl migration of
tricyclic silicon heterocycles that contain exocyclic
chloromethyl substitutents to synthesize silepins (eq 5).9
No migration, however, of a simple alkyl group is observed in


7
this case. Two reports where alkyl migration was demonstrated,
but not examined in detail, are relevant to this study: a)
Voronkov and co-workers have shown low yields of alkyl migra-
tion when potassium fluoride reacts with chloromethyl silanes
in polar solvents at high temperature,10 and b) Hopper and co-
workers have reported methyl migration in hot-tube reactions
over solid potassium methoxide.11
advantages of being anhydrous and of producing highly reactive
"naked" fluoride ion. Typically, solubility problems have
shackled studies of fluoride ion in weakly solvating media.
Crown ethers13 like 18-crown-611 (I) show ionophoric
properties as solid-to-liquid phase transfer agents by com-
plexing metal cations (II) or Lewis acids and so dissolving
metal salts in nonpolar, aprotic media to yield increased
anion reactivity.
Recently, the relative gas-phase acidities of some
(weakly acidic) alkanes (RH) were measured by reaction of
Our reaction conditions12 (eqs 2 and 3) have the
(I)
(II)
1 5
hydroxide ion with trimethyl-substituted silanes.


8
Remarkably, the gas-phase acidities of RH (AHacid) correlate
with the logarithm of the product ratio
(log[Me3SiO /Me2RSiO ]) that results from competitive cleavage
of alkyl or aryl groups and methyl groups from silicon to
produce RH and CH3H (eq 6). Cleavage was demonstrated to be
OH
(CH3)3SiR
(CH3)3SiO + R-H
(6)
(CH3) ,RSiO + CH3-H
controlled by the ability of the migrating group to stabilize
the partial negative charge build-up in the transition state.
In this related solution-phase work, we have used many
of the same R substituents that were used in the gas-phase
study. Also, the range of substituents was extended sys-
tematically to include vinyl and substituted-aryl substituents
of various electron-donating and electron-withdrawing
ability.16 Hammett relative reactivity studies17 were used to
explore the mechanistic detail of the rearrangement (eq 3).
In addition, some other types of chloroalkyl-substituted
silanes have been looked at with a view toward determining the
scope of the rearrangement.


9
2.2 Experimental Results
See the Appendix for general comments on experimental
and instrumental methods.
2.2.1 Reaction of Me3SiCH2Cl with KF
(RD-17-141, SD-1-5, 73)
The following procedure was surreptitiously extracted
from the notebooks of Professor Damrauer. It is included here
as a requisite for this chapter and to show how the work of
this entire monograph began.
A mixture of KF (11.6 g, 200 mmol) and l8-crown-6
(2.50 g, 10.0 mmol) in 10 ml of toluene was prepared in a 50
ml flask furnished with a reflux condenser, a nitrogen inlet
system and a magnetic stirring bar. Me3SiCH2Cl (12.2 g, 100
mmol) was added. The condenser water was maintained at 7 C
using a refrigerated water circulator. Daily GC monitoring
over a two week period showed the build-up of a very short-
retention time peak. At the end of this period a short-path
distillation provided a material boiling between 52-58 C.
Redistillation through a short Vigreux column gave pure
Me2EtSiF boiling at 41-45 C (5.10 g, 48.1$). and 13C NMR
demonstrated the identity of Me2EtSiF and showed that no
Me3SiCH2F was present.
This reaction has also been run on a small scale in a
serum cap vial (SD-1-5, 73) in order to monitor the reaction .
by GC without isolating product(s).


10
2.2.2 Reaction of Me3SiCH2Cl with KI
(SD-1-77)
A mixture of KI (5.80 g, 68.0 mmol) and 18-crown-6
(0.46 g, 1.7 mmol) in 6 ml of toluene was prepared in a serum
cap vial fitted with a magnetic stirring bar. Me3SiCH2Cl
(1.90 g, 15.6 mmol) was added and the reaction vial placed in
a 60 C oil bath. The reaction was monitored by GC for ten
days. Me3SiCH2I was the only product observed to be formed
(by comparison with an authentic sample18), although the
reaction did not go to completion.
2.2.3 Comparison of the Reactivity of Me3SiCH2Cl
with CsF, with KF and with NaF
(SD-1-71, 73, 75, 109, 95)
a) A mixture of CsF (5.15 g, 34.0 mmol), l8-crown-6
(0.45 g, 1.7 mmol), and mesitylene,(0.60 g, 5.0 mmol), as an
internal standard, in 6 ml of toluene was, prepared in a serum
cap vial fitted with a magnetic stirring bar. Me3SiCH2Cl
(1.93 g, 16.0 mmol) was added and the reaction vial placed in
a 60 C oil bath. The reaction was monitored by GC for three
days at which time it was essentially complete. The yield of
Me2Si(F)Et was 87% by the method of response factors. (It
should be noted, as a matter of record, that CsF is extremely
hygroscopic.)
b) A mixture of KF (3.96 g, 68.0 mmol), l8-crown-6
(0.94 g, 3*5 mmol), and meSitylerie (0.88 g, 7.4 mmol), as an
internal standard, in 3 ml of toluene was prepared in a serum


11
cap vial fitted with a magnetic stirring bar. Me3SiCH2Cl
(4.40 g, 36.O mmol) was added and the reaction vial placed in
a.60 C oil bath. The reaction was monitored by GC for two
weeks at which time there was still a substantial amount of
starting silane present. The yield of Me2Si(F)Et was 60-65?
by the method of response factors.
c) Under similar reaction conditions, no reaction of
starting silane was observed to occur with NaF.
2.2.4 Reaction with no l8-Crown-6
(SD-1-103)
CsF (2.58 g, 17.0 mmol) was placed in a serum cap vial
equipped with a magnetic stirring bar. The vial was then
ovendried overnight. Toluene (3 ml) and Me3SiCH2Cl (1.04 g,
8.50 mmol) were added and the reaction vial placed in a 70 C
oil bath. The reaction was monitored every day by GC for one
week. It was very slow to start, but by the third day a
substantial amount of Me2EtSiF had formed. Surprisingly
enough, by the end of the week the reaction was nearly com-
plete.
2.2.5 Reaction with no Solvent
(SD-1-99)
Me3SiCH2Cl (2.09 g, 17.0 mmol) and l8-crown-6 (0.45 g,
1.7 mmol) were added to a serum cap vial fitted with a mag-
netic stirring bar. The crown ether dissolved with stirring.
CsF (2.58 g, 17.0 mmol) was added and the reaction vial placed


12
in a 68 C bath. The reaction was monitored by GC for three
days at which time there was too much salt (presumably CsCl)
and not enough solution to remove anything for a GC. However,
early analysis indicated a reasonable amount of product
(Me2EtSiF) formation after only two hours and >60? completion
on the third day. The reaction was, qualitatively, no faster
without solvent (see 2.2.3). Evidence for another product,
thought to be the result of hydrolysis £(MeaEtSi)20], was also
found by GC.
It should be mentioned that the use of crown ethers
without solvent has been reported.19
2.2.6 Water Content of Oven-dried (120 C) KF
(SD-1-150)
A standard weighing vial was placed in a 600 C oven
and heated to constant Weight (16.3547 g). Prior to all
weighings, this vial was allowed to cool in a dessicator for
15 minutes. KF from our 120 C oven was added to the vial,
which was then weighed (20.5841 g) and placed in the 600 C
oven. The cycle of heating, cooling and weighing was main-
tained until fairly constant weight (20.5798 g) was observed.
The data indicate that 0.0043 g (0.24 mmol) of H20 was lost
from 4.2251 g (73 mmol) of KF or that the formula is
KF*0.003H20.


13
2.2.7 Reaction of Me3SiCHaCl with n-BuN+F~
(SD-1-127, 131, 137)
To a serum cap vial fitted with a magnetic stirring
bar was placed 10.0 mmol n-Bu^F [10.0 ml of a 1.00 M solu-
tion in THF (<5 wt % H20)]. Me3SiCH2Cl (1.04 g, 8.50 mmol)
was added. The vial was capped and the reaction was allowed
to proceed at ambient temperature in one case and at 60 C
(oil bath) in another. Each was monitored by GC for three
days. At the end of this period, in the room temperature
reaction, most of the starting material was consumed. In the
60 C reaction, a substantial amount (about 70$) of the start-
ing material remained. In both cases, Me2EtSiF was the
predominant product. However, the GC analysis was complicated
by a large number of extraneous peaks, many of which result
from the n-Bu3N+F solution itself. At least, one peak is
probably due to the hydrolysis product* (Me2EtSi)20. Further
product analysis was not attempted.
2.2.8 Reaction of Me3SiCH2Cl with KF and MeN(n-octyl)3+Cl
(SD-1-125)
A mixture of KF (0.99 g, 17 mmol) and
MeN(n-octyl)3+Cl (0.34 g, 0.85 mmol) in 1 ml of toluene was
prepared in a serum cap vial fitted with a magnetic stirring
bar. Me3SiCH2Cl (1.04 g, 8.50 mmol) was added and the reac-
tion allowed to proceed at ambient temperature overnight.
Nothing much happened and so the vial was placed in a 60 C
oil bath. Periodic GC monitoring over three days revealed a


14
decrease in starting material and a concomitant increase in
Me2EtSiF. The reaction was reasonably complete (>505?) at this
time. The GC analysis was complicated by a large number of
medium and long-retention time peaks, many of which result
from the tricaprylylmethylammonium chloride. No further
product analysis was tried.
2.2.9 Reaction of Me3SiCH2Cl with KF
in Nonaromatic Solvents
(SD-1-13, 17, 19, 55, 57, 59, 79,
81, 83, 85, 87, 91, 107)
Me3SiCH2Cl, KF and l8-crown-6 (mole ratio = 10:20:1)
were sealed in serum cap vials with appropriate amounts of
these typical crown ether solvents: DMSO, MeCN or DMF.
Pyridine, a very nontypical solvent for crown ethers, was also
tried. In all cases very complex and intractable reaction
mixtures were observed by GC analysis. These were not further
analyzed for product composition.
The same experiments were conducted with Me2PhSiCH2Cl
(see 2.2.12) with the same results.
2.2.10 Reaction using Stoichiometric l8-Crown-6
(SD-1-65)
A mixture of KF (0.27 g, 4.7 mmol) and l8-crown-6
(1.25 g, 4.73 mmol) in 1.5 ml of o-xylene was prepared in a
serum cap vial fitted with a magnetic stirring bar.
Me2PhSiCH2Cl (0.87 g, 4.7 mmol) was added and the reaction
vial placed in a 60 C oil bath. The reaction was monitored


15
periodically by GC for six days. It was reasonably complete
at this time, but was certainly not as rapid as was
anticipated. (See the next section for a detailed description
of the reaction of Me2PhSiCH2Cl with KF.)
2.2.11 Reaction of Me2PhSiCH2Cl with KF
(SD-1-11 21 25, 35, 41 )
A mixture of KF (0.55 g, 10 mmol) and l8-crown-6 (0.13
g, 0.50 mmol) in 0.5 ml of toluene was prepared in a serum cap
vial fitted with a magnetic stirring bar. Me2PhSiCH2Cl (0.87
g, 5.0 mmol) was added and the reaction vial placed in a 60 C
oil bath. The reaction was monitored by GC for three weeks
and at the end of this period structural characterization was
carried out by GC-MS. The MS analysis [m/e (structure, rela-
tive intensity)] is as follows:
Me,Si(F)CH2Ph = Major Primary Product (MW 168)
168 (M+, 36?);
153 (M+ Me, 10?);
77 (M+ CH2Ph, 100?).
MePhSi(F)Et = Minor Primary Product (MW 168)
168 (M+, 48?);
139 (M+ Et, 100?);
91 (M+ Ph, 30?);
77 (M+ MeSi(F)Et, 91?).


16
Me2Si(CHgPh), = Major Secondary Product (MW 240)
240 (M+, 20%);
149 (M+ CH2Ph, 100%);
121 (M+ SiCH2Ph, 46%).
The major secondary product was also identified by GC
and 13C NMR comparison (SD-1-23) with an authentic sample.20
No evidence was found for the other possible secondary
product, MePhEtSiCH2Ph.
2.2.12 Reaction of Me2PhSiCH2Cl with CsF
C SD197)
A mixture of CsF (2.58 g, 17.0 mmol) and l8-crown-6
(0.26 g, 0.97 mmol) in 3 ml of benzene was prepared in a serum
cap vial fitted with a magnetic stirring bar. Me2PhSiCH2Cl
(1.57 g, 8.50 mmol) was added and the reaction vial placed in
a 68 C oil bath. The reaction was monitored by GC for one
week. It was not yet complete at this time, which indicates
no remarkable rate increase in contrast to that observed for
reaction of Me3SiCH2Cl with CsF (see 2.2.3).
2.2.13 Reaction of Me2(CH2=CH)SiCH2Cl with KF
(SD-1-37, 39)
A mixture of KF (0.55 g, 10 mmol) and l8-crown-6 (0.13
g, 0.47 mmol) in 0.5 ml of mesitylene was prepared in a serum
cap vial fitted with a magnetic stirring bar.
Me2(CH2=CH)SiCH2Cl (0.64 g, 4.7 mmol) was added and the reac-
tion vial placed in a 60 C oil bath. The reaction was


17
monitored by GC for three weeks; at the end of this period
structural characterization was carried out by GC-MS. The MS
analysis [m/e (structure, relative intensity)] is as follows:
MegSi(F)CH,(CH=CHg) = Major Primary Product (MW 118)
118 (M+, 36?);
103 (M+ Me, 10?);
77 (M+ CH2CH=CH2, 100?).
Me(CHg=CH)Si(F)Et = Minor Primary Product (MW 118)
118 (M+, 48?);
91 (M+ CH=CH2, 30?);
89 (M+ Et, 100?);
27 (M+ MeSi(F)Et, 91?).
Me2Si(CH2CH=CH2)2 = Major Secondary Product (MW 140)
140 (M+, 20?);
99 (M+ CH2CH=CH2, 100?);
71 (M+ SiCH2CH=CH2, 46?).
No evidence was found for the other possible secondary
product, Me(CH2 =CH)EtSiCH2CH=CH2.
2.2.14 Preparation of Me2(F)SiCH2Cl
(SD-2-81)
To a dry 3-neck 250 ml flask equipped with a magnetic
stirring bar and a condenser was added finely divided SbFa
(17.88 g, 100.0 mmol) under a blanket of N2. This was cooled
to -78 C with dry ice/acetone and Me2(Cl)SiCH2Cl (21.47 g,
150.0 mmol) was rapidly added by syringe. A dry ice/acetone


18
condenser was attached to the water condenser and the reaction
slurry allowed to come to room temperature overnight. The
liquid was then transferred, using a pipet, to a 25 ml flask
and short-path distilled (bp 81 C) to yield Me2(F)SiCH2Cl
(15.95 g, 83.95? yield).21 Structural characterization was
carried out by GC-MS. The MS analysis [ra/e (structure, rela-
tive intensity)] is as follows:
/
Me,Sl(F)CH,Cl (MW 126.65) .
126 (M+, 12?);
111 (M+ Me, 20?);
83 (M+ MeSi, 84?);
77 (M+ CHaCl, 100?);
49 (M+ Me2SiF, 100?);
47 (M+ Me2CH2Cl, 100?).
2.2.15 Reaction of Me2(F)SiCH2Cl with KF
(SD-2-85; SD-3-21)
A mixture of KF (O.073 g, 1.3 mmol) and l8-crown-6
(0.033 g, 0.13 mmol) in 1 ml of toluene was prepared in a
serum cap vial fitted with a magnetic stirring bar.
Me2(F)SiCH2Cl (0.16 g, 1.3 mmol) was added; daily GC monitor-
ing for six days demonstrated the build-up of a single short-
retention time peak. After an additional day at 55 C (oil
bath), no starting material remained. Structural charac-
terization was carried out using GC/MS and demonstrated that
the sole product was MeEtSiF2, the result of Me migration.


19
The MS analysis [m/e (structure, relative intensity)] is as
follows:
MeEtSiF, (MW 110)
110 (M+, 48%);
95 (M+ Me, 84%);
81 (M+ Et, 100%);
67 (not identified, 48%);
47 (M+ MeEtF, 64%).
2.2.16 Reaction of Me3SiCH(Cl)CH3 with KF
(SD-1-47)
A mixture of KF (0.55 g, 10 mmol) and l8-crown-6 (0.13
g, 0.50 mmol) in 0.5 ml of mesitylene was prepared in a serum
cap vial fitted with a magnetic stirring bar. Me3SiCH(Cl)CH3
(0.65 g, 5.0 mmol) was added. The room temperature mixture
was monitored by GC for three weeks. No reaction was ob-
served.


20
2.3 Discussion
We have illustrated that the reaction of fluoride ion
with various chloromethylorganosilanes in nonpolar aprotic
media is unique because of the predominance of the rearrange-
ment pathway to the exclusion of substitution. No substitu-
tion product (e.g., Me3SiCH2F from Me3SiCH2Cl) was ever ob-
tained in these reactions, primarily due to the high fluoride
affinity of the silicon atom.
2.3.1 Reaction of Me3SiCH2Cl with KF and KI
Reaction of Me3SiCH2Cl with KF and a catalytic quan-
tity of l8-crown-6 in toluene exclusively produced the rear-
rangement product, Me2EtSiF, whereas identical reaction with
KI yielded the substitution product, Me3SiCH2I.
The usual order of halide nucleophilicity (a
kinetically-controlled phenomenon) is I > Br > Cl > F ,
which is the inverse of basicity (a thermodynamically-
controlled phenomenon) and the inverse of what would be ex-
pected on steric grounds. However, in apolar nonprotic sol-
vents, the halide ions are free (i.e., not tightly attracted
to a shell of solvent molecules that would obstruct attack of
the substrate) and F is then the strongest nucleoplile and I
the weakest among the halides.
The best leaving groups are the weakest bases. The
more stable it is as a free entity, the more easily a group
comes off. Thus Cl is, typically,, a good leaving group


21
(although I is the best among the halides) and Me is a very
poor one.
Traditionally, it would be anticipated that both F
and I would attack at the carbon of the chloromethyl moiety
because of the presence of a good leaving group and, in fact,
that the reaction with F would be much faster. Why, then,
does F attack the silicon center? The answer may lie in
viewing the initial step of the reaction as a Lewis acid-base
interaction. Tetracoordinate silicon compounds are certainly
Lewis acids since they have vacant (d) orbitals. F is a hard
base and I a soft base.22 We propose a Lewis acid-base
reaction (eq 7) to give a transient negative (pentacoordinate)
F
Me3SiCH2Cl -----
F
Me |
.SiMe ----------- Me2EtSiF (7)
Me | | r\ "Cl"
\J3H2C1
ion in which the silicon center has a higher than normal
valence. A methyl group then migrates, in perhaps the slow
mechanistic step, to displace chloride ion. Of course, sub-
stitution would result if fluoride migrated instead of
methyl. The driving force for the reaction is, apparently,
the immensely strong Si-F bond.
i


22
2.3.2 Various Reagents and Reaction Conditions
These were employed to see a) if the reaction could be
accelerated in order to be synthetically more advantageous and
b) whether the substitution pathway would ever predominate.
Not so surprisingly, the original reagents and reaction condi-
tions (see 2.2.1) proved universally best and no substitution
product ever formed.
The reaction using a stoichiometric amount of 18-
crown-6 (a ten-fold excess over our usual conditions) was
qualitatively faster, but certainly not ten times as fast.
This indicates that the phase-transfer of fluoride into solu-
tion is not the rate-limiting step of the reaction.
Replacing Cl with a better leaving group (e.g., I )
on the a-carbon would probably speed up the reaction, assuming
that organic group migration is the rate-controlling step.
The reagent combination KF/KI could possibly provide an in
situ preparation of the iodomethyl derivative that would then
undergo fluoride-induced migration.9
The data on the water content of oven-dried KF suggest
that our reaction system is reasonably anhydrous, in contrast
to previous work with KF and chloromethylorganosilanes (eq 5).


23
2.3.3 Reaction of Me2RSiCH2Cl (R = Ph and CH2=CH)
In both cases, R migration, to yield Me2Si(F)CH2R,
.dominated by a ratio of 15-20 over Me migration in the early
primary product distribution, as shown by GC analysis. This
suggested that a quantitative treatment (e.g., a Hammett-type
analysis) of the effect of the structure of R on reactivtity
might correlate structural changes with migratory ability.
Over time, the GC peak for Me2Si(F)CH2R was observed
to gradually diminish and the peak for the major secondary
product, Me2Si(CH2R)2, to concomitantly increase.
The mechanistic detail of the secondary reaction as
well as a quantitative treatment of the general rearrangement
reaction has been explored by others (vide infra).16
2.3.4 More Definitive Studies
The real illuminating work on the reaction of fluoride
ion with chloromethylorganosilanes (eq 3) was performed by
other members23* 2<* of our research group and is, perforce,
outside the realm of appropriate discussion here. In terms of
our results, it is sufficient to state that the rearrangement
of Me2RSiCH2Cl with fluoride ion was shown to be controlled by
substituents (i.e., R groups) capable of supporting negative
charge. The R groups that were examined included: methyl,
ethyl, isopropyl, cyclopropyl, butyl, vinyl and phenyl. The
logarithm of the primary product ratio gave good linear cor-
relation with the (Tj. substituent constants of the R groups in


a standard Hammett analysis, which indicated that the ease of
migration was determined by the ability of the migrating group
to stabilize negative charge.
In addition, the primary products of chloromethyl-
dimethyl vinyl and phenyl silanes were found to undergo a
secondary reaction initiated by fluoride ion. To explore the
nature of this secondary reaction, these Ar groups of various
electron-donating and electron-withdrawing ability were
employed in Me2ArSiCH2Cl compounds: p-tolyl, phenyl, p-
fluorophenyl, p-chlorophenyl and m-(trifluoromethyl)phenyl.
2.3.5 Reaction of Me2(F)SiCH2Cl
This reaction was performed in the hope that fluoride
would migrate to yield, in effect, the substitution product;
only the methyl migration product, MeEtSiF2, was, however,
seen. The Taft O'-j- value22 for fluoride (0.51) indicates that
it should migrate much more easily than methyl (-0.05) and
even phenyl (0.10). It must be remembered that values of ffj
are substituent constants for groups substituted at a
saturated carbon and that they represent field effects only.
However, the strength of the Si-F bond is most likely the
determining factor here.


2.3-6 Mo Reaction with Me3SiCH(Cl)CH3
This is probably due to steric hindrance at the o-
carbon and indicates that a fine balance of electronic and
structural effects operate in this reaction. Additionally,
methyl group is a little less electron-withdrawing than a
proton and so the o-carbon would be expected to be a little
less negative than for Me3SiCH2Cl.


CHAPTER III
AN INVESTIGATION OF
. SILACROWN ETHERS
3-1 Introduction and Review
Considerable effort has been devoted to the study of
macrocyclic ligands and their metal complexes since their
remarkable discovery by Pederson in 1967.13 In general, crown
ethers and cryptands have extended coordination chemistry to
the first two groups of the periodic table and, because of the
+ +
biological importance of some of these elements (e.g., Na K
and Caz+), research in this field has been very active. For
example, a large amount of work has been directed toward
elucidating the nature of crown ether complexes and under-
standing the control of the complexation phenomena.25
The silacrowns [generalized structure
R1R2Si(0CH2CH5()n0] have received considerably less attention,
perhaps because they are relatively new. Previous work26 has
indicated that a) they are comparable to simple crown ethers,
like l8-crown-6, in both cation specificity and enhancement of
anion reactivities; b) the Si-OC bond is susceptible to
hydrolysis (particularly at pH extremes); and c) they have
reduced acute oral toxicity compared to the highly toxic
simple crown ethers, which exhibit unusual .long-term pseudo-


27
estrogenic effects even at low levels.
In order to evaluate the phase-transfer ability of the
silacrowns with respect to the fluoride anion, we have used
1,1-dimethylsila-11-crown-4 (III), 1,1-dimethylsila-14-crown-5
(IV) and 1 ,1-dimethylsila-17-crown-6 (V) as phase-transfer
Me Me Me Me Me Me
\ / \ / \ /
Si Si Si
/ \ / \ / \
0 0 ^ r r
0 O'* 0 0 0
w
v_y
III IV V
catalysts for the reaction of F with Me3SiCH2Cl (eq 2).
In general, the reaction is slower than with 18-crown-
6 as the catalyst and a large number of byproducts arise, many
from the silacrowns themselves. In apparent contradiction,
our data show, however, that the silacrowns transfer about 100
times as much fluoride into solution as l8-crown-6. For these
reasons, a study was initiated, using 19F NMR, to determine
the fate of the fluoride ion in solution; specifically, we
wondered if the solubilized fluoride ion attacks the silicon
atom of the silacrown to form a pentacoordinate anion (VI):


28
Me Me
\ /
Si F
/ \
VI
Simply put, our NMR results were indefinite; the
spectra were difficult or impossible to interpret and fre-
quently changed with time. In the hope of "seeing" true
pentacoordinate compounds in solution, we also added Ph2SiF2
or MePhSiF2 to our NMR solutions of silacrown and KF. We
hoped that Ph2SiF3 a well-known pentacoordinate anion, or
MePhSiF3 a heretofore unknown species, might form. Again,
our results were interesting, but inconclusive.


29
3.2 Experimental Results
See the Appendix for general.comments on experimental
and instrumental methods.
3.2.1 Reaction of Me3SiCH2Cl with KF and Silacrown-^
(SD1 113)
A mixture of KF (0.99 g, 17 mmol) and 1,1-
dimethylsila-11-crown-4 (0.18 g, 0.85 mmol) in 1 ml of toluene
was prepared in a serum cap vial fitted with a magnetic stir-
ring bar. Me3SiCH2Cl (1.00 g, 8.50 mmol) was added. The room
temperature reaction mixture was monitored by GC for ten
days. No short-retention time peak ever developed, as would
be expected if Me2Si(F)Et or Me3SiCH2F were produced. Instead,
a number of longer retention time peaks developed. The start-
ing material did not seem to react and the original silacrown-
4 peak decreased in intensity or disappeared. Further product
characterization was considered to be impractical.
3.2.2 Reaction of Me3SiCH2Cl with KF and Silacrown-5
(SD1 115)
A mixture of KF (0.99 g, 17 mmol) and 1,1
dimethylsila-1^-crown-5 (0.21 g, 0.85 mmol) in 1 ml of toluene
was prepared in a serum cap vial fitted with a magnetic stir-
ring bar. Me3SiCH2Cl (1.00 g, 8.50 mmol) was added. The room
temperature reaction mixture was monitored by GC for ten
days. After two days a measurable amount of product having
the same short-retention time as Me2Si(F)Et (by comparison


30
with an authentic sample) was seen; however, most of the
starting material remained unreacted. In addition, the long-
retention time silacrown-5 peak had decreased in intensity and
there were two new medium-retention time peaks and at least
five new long-retention time peaks. After ten days the short-
retention time product peak had disappeared, the starting
material peak was still the peak of greatest intensity and the
medium-retention and long-retention time region was fettered
with peaks. No further product analysis was attempted.
3.2.3 Reaction of Me3SiCH2Cl with KF and Silacrown-6
(SD-1-117)
A mixture of KF (0.99 g, 17 mmol) and 1,1
dimethylsila-17-crown-6 (0.25 g, 0.85 mmol) in 1 ml of toluene
was prepared in a serum cap vial fitted with a magnetic stir-
ring bar. Me3SiCH2Cl (1.00 g, 8.50 mmol) was added. The room
temperature reaction mixture was monitored by GC for ten
days. No short-retention time peak ever developed, which
demonstrates that Me2Si(F)Et was not present or, at least, was
not stable to the conditions of the reaction. The starting
material did not seem to react, but a large number of products
developed. No further product characterization was tried.


31
3.2.4 Reaction of He3SiCH2Cl with CsF and Silacrown-4
(SD-1-119)
A mixture of CsF (2.58 g, 17.0 mmal) and 1,1-
dimethylsila-11-crown-4 (0.18 g, 0.85 mmol) in 1 ml of toluene
was prepared in a serum cap vial fitted with a magnetic stir-
ring bar. Me3SiCH2Cl (1.00 g, 8.50 mmol) was added. The room
temperature reaction mixture was monitored by GC for nine
days. At the end of this period, a small amount (<10%) of
starting material had been converted to Me2Si(F)Et (by com-
parison with an authentic sample) and many other products were
formed. These were not at all similar to those observed for
silacrown-4 and KF (see 3.2.1). Further product characteriza-
tion was not attempted.
3.2.5 Reaction of He3SiCH2Cl with CsF and Silacrown-5
(SD-1-121)
A mixture of CsF (2.58 g, 17.0 mmol) and 1,1
dimethylsila-14-crown-5 (0.21 g, 0.85 mmol) in 1 ml of toluene
was prepared in a serum cap vial fitted with a magnetic stir-
ring bar. Me3SiCH2Cl (1.00 g, 8.50 mmol) was added. The room
temperature reaction mixture was monitored by GC for nine
days. At the end of this period, a reasonable amount (about
50?) of starting material had been converted to Me2Si(F)Et (by
comparison with an authentic sample) and many other products
were formed. These were not similar to those observed for
silacrown-5 and KF (see 3.2.2). Further product characteriza-
tion was not attempted.


32
3.2.6 Reaction of Me3SiCH2Cl with CsF and Silacrown-6
(SD-1-123)
A mixture of CsF (2.58 g, 17.0 mmol) and 1 ,1-
dime thylsila-17-crown-6 (0.25 g, 0.85 mmol) in 1 ml of toluene
was prepared in a serum cap vial fitted with a magnetic stir-
t
ring bar. Me3SiCH2Cl (1.00 g, 8.50 mmol) was added. The room
temperature reaction mixture was monitored by GC for nine
days. At the end of this period, a reasonable amount (40-50$)
of starting material had been converted to Me2Si(F)Et (by
comparison with an authentic sample) and many other products
were formed. The product distribution was similar to, but
cleaner than, that observed for silacrown-5 and CsF (see
3.2.5). The reaction mixture was not further analyzed for
product composition.
3.2.7 Solubility Studies: Silacrowns vs. 18-Crown-6
(SD-1-105)
By stirring KF or CsF with a stoichiometric solution
of crown ether (silacrown-4, silacrown-5, silacrown-6 or 18-
crown-6) in toluene for several days, we have attempted to
quantitatively determine how much fluoride is actually
dissolved. Because this is a very difficult determination in
terms of precision and because most of this work was done by
Professor Damrauer27 in conjunction with Dr. Robert Meglen,28
only our primary operating result will be indicated here. Our
data show, roughly, that the silacrowns transfer 100 times as
much fluoride into solution as l8-crown-6, but that, on a


33
molar basis, <5? of the fluoride is in solution at any one
time.
3.2.8 I9F NMR of KF and 18-Crown-6
(SD-1-69, 149, SD-2-7, 119)
KF (0.29 g, 5.0 mmol) and l8-crown-6 (1.32 g, 5.00
mmol) were combined with 5 ml of toluene in a plastic cap vial
fitted with a magnetic stirring bar. The solution was placed
in a 60 C oil bath overnight. The bright orange solution was
then filtered to removed undissolved KF. l9F NMR, using
external lock, was attempted at ambient probe temperature;
however, no signal could be observed from the solution in an
overnight NMR experiment.29
3.2.9 19F NMR of KF and Silacrown-6 (with Ph2SiF2)
(SD-1-129, SD-2-1)
A mixture of KF (0.29 g, 5.0 mmol) and 1,1
dimethylsila-17-crown-6 (1.47 g, 5.00 mmol) in 5 ml of toluene
was prepared in a plastic cap vial fitted with a magnetic
stirring bar. The solution was allowed to stir for >8 hours.
l9F NMR of an aliquot, using benzene-d6 as an internal lock,
revealed a weak signal at Sp = -134.6 ppm that appears to be a
narrow singlet.
Addition of Ph2SiF2 (0.5 cm) to the NMR tube resulted
in a spectrum with signals at 5p = -133.4 ppm (very intense,
slightly broad) and = -130.0 ppm (very weak, slightly
broad).


3.2.10 I9F NMR of KF and Silacrown-5 (with Ph2SiF2)
(SD-2-3)
A mixture of KF (0.29 g, 5.0 mmol) and 1,1-
dimethylsila-14-crown-5 (1.25 g, 5.00 mmol) in 5 ml of toluene
was prepared in a plastic cap vial fitted with a magnetic
stirring bar. The solution was allowed to stir for >8 hours.
Afterwords, the solution would not settle and so it was
centrifuged at 5000 rpm for a few minutes. l9F NMR, using
benzene-d6 as an internal lock, revealed an extremely weak
signal at 6p = -133.3 ppm that appears be a very complicated
multiplet.
Addition of Ph2SiF2 (6 drops) to the NMR tube resulted
in a spectrum consisting of a single narrow band at Sp = -142
ppm. However, a blow-up of the baseline indicated six addi-
tional signals in the region from 6p = -146 to 130 ppm,
including possible septets (J = 6 Hz) at 5p = -133.4 ppm and
6p = -131.2 ppm.
3.2.11 19F NMR of KF and Silacrown-4 (with
Ph2SiF2 and with MePhSiF2)
(SD-1-139, 143, 147; SD-2-13, 17)
A mixture of KF (0.29 g, 5.0 mmol) and 1 ,1-
dime thyl si la-1 1-crown-4 (1.03 g, 5.00 mmol) in 5 ml of toluene
was prepared in a plastic cap vial fitted with a magnetic
stirring bar. The solution was allowed to stir overnight.
19F NMR, using benzene-d6 as an internal lock, revealed a weak
signal at -134.4 ppm that looks like a singlet. This signal


35
broadens and seems to disappear as the temperature is reduced
from ambient (37 C) to -60 C, although a considerable amount
of variation with temperature was observed for different
solutions at different times.
Addition of Ph2SiF2 (an excess) to the NMR tube
resulted in a spectrum consisting of no less than eight sig-
nals in the region from 6p = -148 to -130 ppm, including
possible septets at 6p = -133.5 and -129.7 ppm and reasonably
broad bands at 6p = -141.9 and -131.1 ppm.
Addition of MePhSiF230 (4 drops) to a separate
KF/silacrown-4 solution gave a complicated spectrum consisting
of septets (6_ = -141.5 ppm, = 133-5 ppm and others of
C ... t
weaker intensity) and broad bands (6p = -136.7 ppm and 6p =
-131.2 ppm). Low temperature spectra (to -35 C) revealed
conversion of the broad bands to multiplets (a quartet and a
septet, respectively) which is indicative of an exchange
process at ambient probe temperature.
3.2.12 I9F NMR of Ph2SiF2 and MePhSiF*
(SD-1-147b; SD-2-13)
A solution of Ph2SiF2 'in toluene with benzene-d6 as an
internal lock provided a narrow singlet at 6p = -142 ppm with
evidence of F-Si coupling (Jpgi = 2^ Hz^*
A solution of MePhSiF2 in benzene-d6 yielded a quartet
(JFSiCH = 6 Hz) at 6F = "137 PPm With JFSi = 290 Hz*


36
3-3 Discussion
The silacrown ethers have proven to be highly complex
in their interaction with fluoride ion, although they have
been shown to be advantageous26 as phase-transfer catalysts
for the potassium salts of other ions (e.g., CN ). The source
of this complexity may reside in the attack of F on the Si
atom of the silacrown.
3.3.1 Reaction of Me3SiCH2Cl with F and Silacrown
The best combination of reagents seems to be
silacrown-5 (IV) or silacrown-6 (V) with CsF. However, with
all reagents, the reaction is slow and a large number of
products are produced. The silacrowns are certainly destroyed
as shown by the large number of long-retention time peaks in
the GC; perhaps, ring-opening occurs upon fluoride'attack at
the sila center of the crown ether.
3.3.2 I9f nmr
The region in which all of our chemical shifts occur
Up = -130 to -146 ppm) is certainly suggestive of fluorine
r
bound to silicon (see Chapter IV).
For solutions of KF and silacrown, the fluorine
resonances are so weak that F-Si coupling, if it is present,
cannot be observed. A solution of KF and l8-crown-6 gave no
fluorine signal (but it can be argued that we did not try hard
enough to find onee.g., we did not look at the "precipitate"


37
of this reaction). Therefore, it is impossible to know
whether these fluorine resonances arise from "naked" or bound
fluoride ion, although we favor the latter. The most promis-
ing solution for further study is the KF/silacrown-4 mixture;
and, CsF/silacrown solutions should be looked at.
When either Ph2SiF2 or MePhSiF2 is added to mixtures
of KF and silacrown, highly challenging, but as yet inex-
plicable, phenomena are observed. The silacrown-4 solutions
are particularly interesting because of the presence of a)
splitting (i.e., septets that reflect coupling of F to six Me
hydrogens?) and b) temperature-dependent behavior (indicative
of pentacoordination?). By their fluorine chemical shifts, it
is clear that Ph2SiF2 and MePhSiF2 do not remain unreacted.
However, the abundance of signals argues that many different
species are presentand that is discouraging. More careful
studies though are called for.


CHAPTER IV
THE REACTION OF FLUORIDE ION WITH ORGANOFLUORSILANES
TO FORM PENTACOORDINATE SILICON ANIONS
1 Introduction and Review
A recent review31 makes it apparent that or-
ganofluorosilanes (compounds containing the C-Si-F linkage)
have a rich and varied chemistry. The small size of the
fluorine atom and the extremely strong Si-F bond account, in
part, for the many unique features of these compounds. In
addition, organofluorosilanes are particularly amenable to
multinuclear NMR studies because a) fluorine is the third most
sensitive NMR nucleus; b) fluorine-silicon coupling, which
results from the 29Si isotope (relative abundance 4.67?), is
readily observable; and c) these coupling constants are typi-
cally large,.viz., > 150 Hz.
Examples of pentacoordination pervade the periodic
table. In addition, many proposed reaction mechanisms invoke
a transitory hypervalent intermediate.32 In particular,
reaction pathways which involve the extention of coordination
at silicon are quite prominent.33 Therefore, it seemed ap-
propriate to focus attention on pentacoordinate silicon
species themselves.


39
We have discovered a new method31* for making pen-
tacoordinate silicon anions (i.e., siliconates) by reacting
suitably-substituted organofluorosilanes with stoichiometric
amounts of KF and l8-crown-6 (eq 8). The purpose of this
D 1 D 2d 3c-? I? KF 18-crown-6 . ro F ^ L .SiR3 (8) R1 1 F
n it u oil* toluene
R1 R2 R3 Compound
Ph Ph F VII
Me Ph F VIII
F Ph F IX
Ph Ph Ph X
m-FPh Ph m-FPh XI
m-FPh m-FPh m-FPh XII
o-xylyl o-xylyl F XIII
F o-xylyl F XIV
chapter is to discuss the low-temperature dynamic multinuclear
NMR behavior of the siliconates we have characterized (VII -
XIV ) detailing a) proposed geometries, b) activation barriers
for the dynamic process(es), c) the comparison of such bar-
riers to those already measured in isoelectronic, isostruc-
tural phosphorus compounds and in cyclic siliconates, d)
whether different mechanisms of dynamic exchange operate at


40
different temperatures, e) the effect of concentration, sol-
vent polarity and/or inherent impurities on exchange rates or
mechanisms of exchange, and f) evidence for propeller-like
conformations in certain multi-aryl siliconates (e.g., XIII).
In addition, work we have just initiated on an unusual dis-
proportion reaction (compound XIII ----- compound XIV) will be
briefly considered. Finally, we have done some exploration on
the scope of eq 8 and, so, present information on the types of
potential siliconates that give atypical results, or that we
cannot form.
4.1.1 Pentacoordinate SiliconEarly Work
The two common pentacoordinate geometries, the
trigonal-bipyramidal (TBP) and the rectangular pyramidal (RP),
L L
Lv! 'mL LA
\St
L L
TBP ]
M L
RP
are unique among geometries encountered for structures with
simple coordination numbers because a nonequivalence of posi-
tion exists. This results in two sets of bonds with con-
siderably different properties and two distinct types of
bonding ligands, i.e., apical and equatorial groups for the
TBP geometry and apical and basal groups for the RP geometry.


41
The RP form seems to predominate for transition metal ele-
ments, while for nontransition elements the main conformation
for pentacoordinate compounds is TBP. However, a range of
structural forms, which would include the entire spectrum of
intermediate geometries, is certainly to be expected, even for
M = Si.35 Both experimental and theoretical work36 suggest
that, for M = nonmetallic, the TBP geometry is preferred by
compounds which have two electronegative groups (apical) and
three electropositive groups (equatorial), whereas a molecule
with four electronegative groups will be distorted toward the
RP form with these groups in basal sites.
Previous work37* 38< 39 with fluorophosphoranes (e.g.,
Ph2PF3), using low-temperature 19F NMR as the main
stereochemical probe, has revealed several generalities con-
cerning the TBP structure: a) apical bond lengths are typi-
cally greater than equatorial bond lengths, using fluorine
coupling constants to phosphorus as evidence; b) the more
electronegative and weakly ir-bonding groups tend to assume
apical positions; c) the ligands are susceptible to fast
intramolecular positional exchange, with Berry pseudorotation
(vide infra) as the most generally accepted mechanism of
intramolecular apical-equatorial interchange; and d)
equatorial fluorine atoms are shielded (i.e., upfield) and
apical fluorines are deshielded (i.e., downfield) relative to
the averaged fluorine signal in the l9F NMR spectrum.


42
Pseudorotation, as a mechanism for intramolecular
rearrangements, refers, in general, to any unimolecular
process internal to a molecule for permuting the positions of
groups about a central atom. The Berry mechanism1*0 is a
pseudorotation process for simultaneously interchanging two of
the equatorial groups with the two apical groups, while the
third equatorial group (the pivot) remains equatorial (eq 9).
1
I
M3
I
2

+
TBP
RP
TBP
(9)
The transition state is a rectangular pyramid. This
mechanism, or some variant thereof, is often invoked to ac-
count for the stereochemical results of nucleophilic substitu-
tion at chiral, tetrahedral silicon;, groups have been presumed
to enter and leave the intermediate (anionic) structure from
apical positions33 and from equatorial positions.1*1 In addi-
tion, recent X-ray work on siliconate geometries implies that
pseudorotational processes, already well-established for
phosphoranes, should exist for pentacoordinate silicon;
however, the barriers to ligand exchange (i.e., AG*) are
expected to be not quite as large.38


^3
The study of pentacoordinate organofluorosilicon
anions in solution began with the pioneering endeavors of
Klanberg and Muetterties (K-M) in 1967 -**2 These authors
prepared the anions XV XVIII as their tetra-alkyl ammonium
salts in an exothermic reaction (eq 10) and reported on the
R1R2SiF2 (n-C3H7)N+F~ (n-C3H7)N+ F rn I- .SiF (10) R1 | F
MeOH
R1 R2 Compound mp (C)
Ph Ph XV 65-67
F Ph XVI 102-103
F Me XVII 161-;:^
F F XVIII 205-207
low-temperature dynamic 19F NMR behavior of these species in
CH2C12 (XV, XVII and XVIII) . in CHC13 (XVI) and in MeOH
(XVII). The composition of the salts did not depend on the
molar ratio of the reactants used. Klanberg amd Muetterties
were not able to make Ph3SiF2 (IX), Me2SiF3 or Me3SiF2 but
did prepare two hexacoordinate, complexed di-anions, which
were synthesized (eqs 11 and 12) and studied in H20.



SiF
NHF
Ha0
- (NH )
i* / 2
F I F
r V. 1 S
^Si
FX | VF
F
(IXX)
NHF (11)
MeSiClg
or
MeHSiCl2
NHUF
H-0
-* (NH )2
Si,
Me-" |
(XX)
NHF (12)
(Of historical interest is the citation1*3 that Berzelius
prepared SiFs2 salts from SiF,, and metal fluorides in 1828.)
When K-M investigated SiF5 (XVIII), they found
spectroscopic equivalence of fluorine atoms (6p = -136.0 ppm)
throughout the temperature range of investigation (to below
-60 C). This was not surprising because all previously-known
monomolecular MFS species exhibited fluorine equivalence and
had been shown to have TBP geometry.37 At higher tempera-
tures, the fluorine resonance began to broaden and no F-Si
coupling could be observed. However, at < -60 C, spin-spin
coupling was seen and it was noted that the coupling constant
(JFS1 1,8 Hz> was approximately the arithmetic mean of the
values for SiF,, [J. = 178 Hz (S_ = -160.3 ppm)] and SiFsz
r ol r
[JpSi = 110 Hz (6^ = -128.0 ppm)], conforming with the ap-
parent decrease of s character in the Si-F bonds from SiF,, to


45
SiFfi
These results were interpreted to mean that at tem-
peratures below -60 C an intramolecular rearrangement (prob-
ably Berry pseudorotation) was the dominant exchange process
for the TBP anion; however, at temperatures above -60 C,
another dynamic mechanism, which must involve Si-F bond
cleavage, was occurring. K-M conceded that trace impurities
might catalyze Si-F bond-breaking and, in fact, discussed the
considerable difficulties posed by impurities:
The crude products invariably contain small but
highly detrimental amounts of impurities, detrimental in
the sense that these impurities (usually hydrolysis
products) grossly perturb the 19F NMR resonances.
Painstaking purification from nonprotonic solvents is
usually necessary to obtain products of sufficient purity
for a meaningful NMR investigation.
Nevertheless, they felt that scission of the Si-F bonds could
also result from a dissociative-type process (eq 13) or an
SiFs t SiF + F (13)
associative one leading to fluorine-bridged dimers (eq 14),
2SiFs
F F F F F
\l/ \l/
Si Si
/|\ /l\
F F F F F
(14)
which they favored since XVIII should have significant ac-


46
ceptor properties. K-M believed that XVIII had TBP geometry,
by analogy with the structural principles previously estab-
lished for other MF5 compounds.
Similarly, PhSiF (XVI) and MeSiF~ (XVII)
demonstrated spectroscopic equivalence of fluorine atoms at
all temperatures (to -125 C). Compound XVI gave a narrow
peak with spin-spin coupling [Jpgi = 206 Hz (5? = -119.2 ppm)]
at room temperature; no change occurred at lower
temperatures. This data did not permit the assignment of a
stereochemistry, but K-M proposed TBP geometry with the phenyl
group in an equatorial position. The 1F signal for XVII at
room temperature was a broad doublet, apparently as a result
of exchange phenomena. At -60 C, the peak had narrowed to a
quartet (coupling of the three protons to four equivalent
fluorines) and silicon satellites were observed [JFSi = 218
Hz, JFSiCH = 4.8 Hz (5p = -110.9 ppm)], indicating no rapid
Si-F bond breaking. Consistently, the methyl resonance in the
*H spectrum was split into an AB,, quintet. Again, a TBP
geometry with the methyl group in an equatorial position was
proposed, by analogy with monoalkylfluorophosphoranes. K-M
favored a dissociative exchange process (eq 13) at higher
temperatures for XVII; apparently, no intermolecular exchange
was occuring for XVI.
Additionally, K-M noted that when nonpolar solutions
of RSiF (R = Ph or Me) were subjected to fluoride ion or a


47
protonic substance (MeOH or H20) there resulted: a) upfield
shifts (ca.,10 ppm) and broadening (to 100 Hz) of the I9F
resonances and b) loss of silicon satellites. They felt that
exchange, due to a bimolecular process (eq 15), was occurring.
RSiF + F*~ t RSiF3F*~ + F~ (15)
The data for Ph2SiF3 (XV) were much more interesting
from a stereochemical point of view. The 19F NMR spectrum was
a single broad peak [SF = -110 ppm (BWHM = 70 Hz)] at ambient
temperature, which showed no evidence of F-Si coupling. From
-20 to -60 C, the bandwidth was so large as to render the
resonance undetectable. Below -60 C, two peaks, characteris-
tic of nonequivalent fluorine atoms, appeared. They were
initially quite broad, but sharpened somewhat as the tempera-
ture approached -100 C, whereupon silicon satellites appeared
CJF(eq)si = 212 Hz ( JF(ax)Si = 25>i iiz (6F(ax) = "98- PPm integral = 2)]. The
difference in the chemical shifts of the nonequivalent
fluorines was quite large (A = 36.0 ppm), and the weighted
average of these shifts [(-134 + 2(-98))/3 = -11Q] was consis-
tent with the ambient chemical shift. No F-F coupling was
observed, but K-M considered this a reflection of the intrin-
sically broad resonances observed at -100 C and the estimated
small coupling (<5 Hz). By analogy with Ph2PF3, the data were


48
considered consistent with a pentacoordinate species of TBP
geometry with the phenyl substituents in equatorial positions.
Because of the temperature data, XV was interpreted to be the
least stable toward intermolecular exchange. Thus, K-M
favored a dissociative-type process' (eq 13) as the dominant
feature of this system.
It is interesting to combine these results with the
predictions of K-M and with more recent 29Si NMR data. In
extrapolating from the data obtained from XV, K-M predicted
that in solution the ion Me2SiF3 if made, would be largely
dissociated and undergo very rapid intermolecular fluorine
exchange, even at low temperatures. Me3SiF2 would be the
least stable of the alkyl-substituted ions. (No prediction
was made for Ph3SiF2 .1*1*) K-M expected that the replacement
of fluorine ligands with alkyl substituents lowered Si-F bond
energies, but that this happened to a lesser extent with
aromatic ligands. Thus, they hypothesized that the order of
acceptor ability is: SiFs > PhSiF,, > MeSiF,, > Ph2SiF3 >
Me2SiF3 > Me3SiF2 [Apparently, PhSiF,, is on the border
between associative-type processes (eq 13) and dissociative-
type processes (eq 12).] This order is in accord with the
accepter ability of the four-coordinate starting materials
(when viewed as Lewis acids).
The 29Si chemical shifts1*5 of the four-coordinate
starting materialsincluding Ph3SiF and MePhSiF2, with which


49
we were able to make pentacoordinate speciesgive an identi-
cal order (Table II), although the 19F chemical shifts and the
Table II. NMR data of four-coordinate starting silanes
Compound 29Si 6 (ppm) 19F 6 (ppm) J (Hz)
SiF -113.7 -160.0 170
PhSiFj -73.2 -143.0 268
MeSiF3 -56.2 -139.2 267
Ph2SiF2 -30.5 -144.1 290
MePhSiF2 -12.4 -137.0 287
PhjSiF -4.7 -170.0 281
Me2SiF2 4.4 -133.3 288
Me3SiF 30.5 -159.3 276
silicon-fluorine coupling constants (J) do not correlate in a
similar manner.
4.1.2 Pentacoordinate SiliconRecent Work
Recent NMR studies of the dynamic behavior of pen-
tacoordinate silicon have concentrated on a) the nature of
extraneous (i.e., non-pseudorotational) ligand exchange
processes in solution which may involve impurities (e.g., H20
and HF), or solvent or other silicon species and on b) exotic,
cyclic silicon compounds, many of which are not, necessarily,
organofluorosilanes.


50
Gibson, Ibbott and Janzen1*6 (1973) have presented
evidence that intermolecular fluorine exchange in SiF5
(XVIII) is due to hydrolysis (eq 16).. They added
SiFs + H20 t SiF0H + HF (16)
hexamethyldisilazane1*7 to an NMR sample of XVIII and detected
coupling (JFSi = 1^0 Hz) at,38 C, which indicates that inter-
molecular exchange had been stopped or, at least, slowed
down.1,8 Introduction of H20 resulted in the expected loss of
F-Si coupling. These authors recognized that a lower energy
intramolecular mechanism of exchange was still equilibrating
apical and equatorial fluorines since they observed fluorine
spectroscopic equivalence to -90 C. They proposed that trace
amounts of Lewis bases (e.g., H20 or NH3) may coordinate with
XVIII in solution to yield a rapid interconversion between
five and six coordinate geometries (eq 17). The result is
SiFs + NH3 t SiF5:NH3
(17)
that, at all temperatures, the positions of the fluorine atoms
are averaged, but F-Si coupling is retained. This process
was, therefore, regarded as an intramolecular ligand exchange
due to Lewis acid-Lewis base interactions. Although they are
not clear about this, the authors seem to view their proposal


51
as an alternative mechanism, occurring in the presence of
impurities, to pseudorotation.
In 1977, Marat and Janzen1*9 investigated catalyzed
intermolecular fluorine exchange in MeSiF,, (XVII) and con-
cluded that it was a very complex process. Their results
indicated that negligible intermolecular exchange occurred
unless a catalyst (or impurity) was added. Therefore,
unimolecular ionization of XVII (eq 18), or bimolecular
MeSiF,, t MeSiF3 + F~ (18)
fluorine exchange of XVII (eq 19), could not be mechanisms
MeSiF,, + MeSiF*,, t MeSiF3F* + MeSiF*3F_ (19)
of intermolecular exchange. HF, H20 and MeOH demonstrated
catalytic ability; hexamethyldisilazane, pyridine, F and
MeSiF,, were inhibitors of fluorine intermolecular exchange.
They concluded that HF was the dominant fluorine transfer
reagent and that H20 (and MeOH) react with XVII to produce HF
(eq 20):
MeSiF,, + H20 t MeSiF30H + HF (20)


52
Marat and Janzen50 also examined intermolecular
fluorine exchange in the systems SiF-SiFs MeSiF3-MeSiFH ,
PhSiF3-PhSiF,, and in SiFs -SiF62 SiF,,-SiF62 and
MeSiF3-SiF62 They observed rapid fluorine exchange (i.e.f a
single, slightly broad, averaged peak in the 19F NMR spectrum)
between four- and five-coordinate silicon fluorides and ex-
plained this result in terms of a rapid equilibrium to form an
intermediate containing a fluorine-bridged bond (eq 21). A
RSiF3 + RSiF
F F
\ /
FSiF
R
I
SiF
I / \
R F F
(R = F, Me, Ph)
(21)
similar interpretation was given for the observed exchange
phenomenon between five- and six-coordinate silicon. Inter-
estingly enough, they also showed that, for the Si(4)Si(6)
systems that were examined, the formation of pentacoordinate
silicon anions was thermodynamically and kinetically favored
(eq 22):
RSiF3 + SiFs2 t RSiF + SiFs (22)
(R = F, Me)


53
Marat and Janzen concede that the previous work
provides no direct evidence related to the mechanism of
apical-equatorial (i.e., intramolecular) fluorine exchange.
However, they postulate that exchange occurs in an equilibrium
process via either or both of the intermediates shown below
(eq 23):
*V!-
SiF -
I
SiF.
Base or
Solvent
\ / 1
Si FSi
1 F^ I
F F
Fv I F
^Si
FX | VF
Base
(23)
Corriu and co-workers51 (1980 ff.) have'chosen to
study neutral TBP silicon compounds in which pentacoordination
results from intramolecular N*Si coordination (e.g., compounds
o
and XXIIb). These systems were chosen because
Y2 F F
1 /Me Si t^Y1 -IC l/F Si tXF
-NMe i N NMe ^ I NMe |
1 Me 1 Me Me
XXI
XXIIa
XXIIb


54
a) they help to avoid problems inherent with impurities and b)
they are seen as models of nucleophilic displacement at
silicon.
In XXI, a variety of Y groups (vide infra) were used;
some of the neutral compounds became pentacoordinate at low
temperatures and others did not. The formation of an in-
tramolecular N-*Si bond was detected by the resultant dias-
tereotopy of the Me groups on nitrogen (due to chirality at
silicon). Interestingly, coordination was not a function of
the electronegativity of Y, but was controlled by the ability
of the nitrogen atom to stretch the Si-Y bond. Corriu's data
led to the following order of increasing pentacoordination
ability for the these Y groups (eq 24):
H, OR < F, SR < OAc, Cl, Br (24)
In the low temperature l9F NMR spectrum of XXIIa and b
in non-polar media, apical and equatorial fluorines were
observed. At ambient temperature, only a single fluorine
resonance was seen and F-Si coupling was present throughout
the temperature range, which indicated that a dynamic in-
tramolecular process was responsible for making the fluorines
equivalent at room temperature. Corriu postulated that,
because of the lability of coordinative bonding, an irregular
mechanism, consisting of fission, rotational movement of the


55
intermediate tetragonal silicon and reformation of the N+Si
bond, best described the equilibration process.
It is interesting to examine Corriu's results for
F -F coupling, although curiously they are not directly
a 6
addressed in these papers. For XXIIa (coalescence tempera-
ture, Tq est. -HO C), the fluorine-fluorine coupling was of
medium magnitude (J-.-p = 15 Hz) at -92 C and the difference
in apical/equatorial shifts was, not unexpectedly, quite large
(A = H3.0 ppm). In XXIIb (T est. -5 C), substantial cou-
pling was present (JpSiF = 1+2 c but the dif-
ference between the apical doublet and the equatorial triplet
was quite small (A = 20.0 ppm).
A new class of neutral TBP silicon compounds has
recently been reported by Voronkov and co-workers (1980).sz53
The (aroyloxymethyl)trifluorosilanes (e.g., compound XXIII)
F
p-CH
H,CSi
0'
,0C
; flu
,C=0
F
F
XXIII
resemble the Corriu compounds in that they have a nucleophilic
atom in the 6-position of the side chain and so can form a
five-membered ring structure with silicon to give
pentacoordination. While two fluorine signals were seen at


56
low temperature for XXIII, only one signal, with 29Si satel-
lites, was observed in the 19F NMR spectra at ambient
temperature. This, of course, indicated that a rapid (on the
NMR time scale) F(a)F(e) intramolecular exchange was
occurring. The authors did not establish whether the exchange
was due to silicon pseudorotation and/or successive 0*Si
cleavage and bonding.
In accord with the results of Corriu, the fluorine-
fluorine coupling for XXIII (TQ = -60 C) was large (JpSip =
47 Hz), while the difference between the apical and equatorial
shifts was quite small (A = 14.2 ppm). In addition, the F
d
chemical shifts for compounds like XXIII with different
aromatic substituents were linearly dependent on the Hammett
substituent constants, However, the relationship between the calculated AG* values of
these compounds and the Hammett constants was nonlinear.
Farnham and Harlow51* (1981) have prepared two spiro
siliconates (compounds XXIV and XXV) in acetonitrile using
XXIV
XXV


57
tris-(dimethylamino)sulfonium (TAS+ = M+) trimethyldifluoro-
siliconate55 as the fluoride source.
Both siliconates exhibited temperature-dependent I9F
NMR spectra. The silicon-bound fluorine of XXIV did not
undergo intermolecular exchange at a significant rate as
confirmed by the 29Si NMR spectrum [5gi = -76.6 ppm (doublet,
Jg^p = 227 Hz)]. Visual fit of observed and simulated spectra
at various temperatures gave AG* = 16.6 kcal/mol (AH* = 14.1
kcal/mol; AS* = -8 eu). The authors considered the Berry
pseudorotation process to be the most reasonable enantiomeric
mechanism, although they could not exclude ligand permutation
by Si-0 bond breaking. In contrast, there was no spin cou-
pling of the unique fluorine in XXV [5gi = -75.1 ppm
(singlet)], even in the presence of added.HMDS, revealing that
this fluorine ligand was undergoing very rapid intermolecular
exchange. And, a plot of In k vs. 1/T was nonlinear, which
was interpreted to be indicative of a change in mechanism with
temperature. The authors felt that many alternative
mechanisms were possible for XXV (e.g., HF-catalyzed apical-
equatorial fluorine exchange) and that some might act in
concert with pseudorotation.
Stevenson, Wilson, Martin and Farnham56 (1985) have
'extended studies on anionic spiro bicyclic siliconates by
preparing the compounds XXVIa-j (Table III), by addition of


58
Table III. Spiro siliconates
Compound Y nucleophile Y to the distorted-tetrahedral starting silane;
the rates of ligand permutation (i.e., inversion of configura-
tion at silicon) were ascertained via 19F NMR monitoring of
site interchange for the geminal CF3 groups.
The data were determined to be consistent with a
mechanism proceeding by nondissociative, Berry-type
pseudorotation. The observations of a small entropy of ac-
tivation AS0 [-10.9 eu (XXVIb), 4.5 eu (XXVIg), -8 eu (XXVIi)
and 3.2 eu (XXVIj)] and a neglible solvent polarity effect
were interpreted as providing support for an intramolecular
process of ligand permutation as opposed to a dissociative
mechanism involving, e.g., Si-0 heterolysis. In addition,


59
there was excellent linear correlation between the energy
barriers AG (kcal/mol) at 151 C to inversion of silicon and
the Taft = -3.37 kcal/mol)despite the use of various counterions,
solvents and measurement methods. The values of AG decreased
(vide supra) as the electron-withdrawing inductive effect of
the substituent Y increased. In other words, electron-
withdrawing (i.e., apicophilic or (^acceptor) equatorial
substituents favored pseudorotation.
The determinations of AG for inversion by pseudorota-
tion allowed the first direct comparison of energy barriers to
pseudorotation between isoelectric, isostructural derivatives
of nonmetals of different columns in the periodic table. The
siliconate XXVId has a lower barrier to inversion (26.0
kcal/mol) than.does the equivalent phosphorane analogue (28.3
kcal/mol). The greater diameter of anionic silicon relative
to neutral phosphorus was thought to lower the barrier to
pseudorotat ion.5 7
Notable among the spectroscopic evidence for the
proposed pentacoordinate structure of these siliconates was an
upfield 29Si shift (6Si = -64.1 to -82.4 ppm) relative to the
starting silane (5Si = 8.6 ppm). This has been shown to be
indicative of an increase in the coordination state of
silicon.38 Also, the observed low-field chemical shift (5H =
8.05-8.40 ppm) of the ortho-aromatic protons has been shown to


60
be a special feature that is characteristic of such protons in
many analogous TBP derivatives of other hypervalent nonmetals
(e.g., P, S and I).59 This was interpreted to result from the
polarization of the ortho-aromatic C-H bond, as a consequence
of its proximity to the highly polar Si-0 bond.
Although most of the siliconates XXVI were stable to
the above-ambient temperatures required in this study, at
temperatures > 110 C one unusual decomposition slowly oc-
curred wherein the Y = pentafluorophenyl compound became
converted to the Y = fluoride compound (eq 25). At 150 C,
XXVI g
(25)
19F NMR indicated 52? conversion after 100 minutes. No decom-
position was observed in the presence of excess HMDS,1*7 which
suggested that a fluoride ion chain was involved. The source
of the fluoride ion was postulated to be the pentafluorophenyl
anion, which could lose F to become a benzyne-like neutral
species. The decomposition of XXVIg to give pentafluorophenyl
anion (and starting silane) was thought either to be a


61
unimolecular decomposition or promoted by adventitious
nucleophiles. It is important to note that 23? of an uniden-
tified product was also present in the 19F NMR.
It should now be obvious that a great deal of atten-
tion must be paid to the ligands on a tetrahedral silicon if
one hopes to make a pentacoordinate species that is stable in
solution. Aromatic groups, electronegative groups and dis-
torted tetrahedral structures seem to favor the formation of
isolable pentacoordinate species. In particular, alkyl sub-
stituents on silicon do not facilitate the construction of
hypervalent compounds that can be studied by dynamic NMR
techniques. However, it is worth noting that compound XXVII,
XXVII
a fully alkyl carbon-substituted siliconate, has been prepared
in the gas phase (Damrauer et al, 1981).60
4.1.3 Propeller Compounds
The most important and clarifying work in the field of
molecular propellers has been performed by Mislow et al.61
This group has been interested in the restricted internal
rotation and the correlated motions of triarylboranes (Ar3B),


62
triarylmethanes (Ar3CH), diaryl systems [ Ar2ZXY (Ar2CHCl,
Ar2PH); Ar2ZX2 (Ar2CH2, Ar20)], tetraarylmethanes (Ar,,C) and
tetrarylethanes [(Ar2CH)2]. It is neither necessary nor
possible to review these studies here; however, the results
conveyed in many lengthy papers62 for, e.g., triaryl deriva-
tives of the type Ar3CH, are overwhelmingly in favor of the
chiral (C3) propeller conformation in solution, where the
aromatic rings are twisted about their local C2 axis like the
blades of a three-bladed propeller. Several mechanisms have
been considered for the transformation of one enantiomer into
the other via correlated ring rotations;63 compelling
theoretical and experimental evidence has been presented for
the operation of the so-called two-ring flip mechanism, whose
transition state, looking along the C-H bond, is indicated
schematically in (XXVIII), as the stereoisomerization
XXVIII
pathway of lowest energy. In this mechanism, two aryl rings
rotate through planes defined by their local C2 axes and the
methine C-H bond, and the remaining ring passes through a
perpendicular plane.


63
The details of the geometric structure of the
triphenylmethyl (trityl) cation remained a challenging problem
for many years. Lewis and Calvin61* (1939) were the first to
point out that the molecule must necessarily assume a confor-
mation that is a compromise between minimum steric repulsion
and maximum resonance stabilization since severe steric inter-
actions between ortho hydrogens of neighboring phenyl rings-
presumably prevented the coplanar geometry required for maxi-
mum positive charge delocalization. The chiral propeller,
which requires the presence of two enantiomeric forms (XXIXa,
and XXIXb), became the most generally accepted conformation.
Schuster, Colter and Kurland63 (1968) provided the
first quantitative measure of the conformational stability of
trityl cations by examining the low-temperature 19F NMR
spectra of (m-FPh)3C+ (XXX) and (m-FPh)2PhC+ (XXXI). They
recognized that, because the enantiomers differed in the pitch
of their phenyl rings, interconversion could occur by rotating
each phenyl ring either through or perpendicular to the
trigonal plane. Because the first motion would lead to in-


64
creased steric repulsion and the second to decreased charge
delocalization, it was important to study the mechanism and
energetics for the interconversion to decide whether steric or
electronic effects dominated. Both compounds gave a single
fluorine resonance at ambient temperature under proton decou-
pling conditions (Sp = -111.8 ppm and 5p = -112.5 ppm,
respectively). (The upfield shift from XXX to XXXI was inter-
preted to reflect increasing electron donating ability accord-
ing to the order: m-FPh < Ph < p-FPh.) The signal for XXX
broadened at -30 C, split into two lines and remained un-
changed from -51 to -80 C. The resonance for XXXI broadened
at -38 C and became a four-line pattern that remained un-
changed from -69 to -80 C. The authors concluded that the
barriers to interconversion were ca. 14 kcal/mol and that
isomerization occurred by a three-ring flip transition state.
Rakshys, McKinley and Freedman65 later demonstrated that the
two-ring flip mechanism was actually favored.
We are aware of only one previous effort to study
isomerization in propeller-like organofluorosilanes. Boet-
tcher, Gust and Mislow66 (1973) recorded the variable-
temperature *H NMR spectra of (o-xylyl)3SiF (XXXII) as a 10£
solution in CS2. It was anticipated that, in the absence of
rotation about the C-Si bonds, the two diastereotopic methyl
groups would yield two separate signals in the *H NMR spectrum
(i.e., in a propeller conformation, each aryl group would have


65
a methyl group proximal to the Si-F bond and one distal). The
ambient temperature spectrum demonstrated a doublet in the Me
region [5H = 2.12 ppm (J^ccCSiF = 2*^ Hz^* As the tempera-
ture was decreased, this signal broadened (T = -2H C),
v
disappeared, and reappeared (T = -76 C) as a doublet =
2.25 ppm (JHQCCSiF = 5.5 Hz)] and a singlet (5R = 1.83 ppm).
The spectra were simulated using the computer program DNMR267
to allow calculation of the rate constants k for exchange of
magnetic environments of the ortho methyl groups. The rates
obtained were used to determine a free energy activation
barrier of AG^Q = 12.1 kcal/mol for enantiomerization, as
calculated from the Erying equation60 (eq 26). It was
k = k(kgT/h)exp(-AG^/RT) (26)
assumed that the two-ring flip mechanism operated and so the
rate of enantiomerization equalled 3kQ. The authors concluded
that barriers to enantiomerization in triarylsilanes were much
too low to permit facile resolution.


66
4.2 Experimental Results
See the Appendix for general comments on experimental
and instrumental methods.
4.2.1 NMR Data for Ph2SiF2
(NMR-1a)
The spectra for a commercial sample were recorded as:
*H NMR 5 [multiplicity]
7.30-7.90 Cm].
19F NMR 6 [multiplicity]
-141 .4 [s (JFS. = 291 Hz)].
4.2.2 Reaction of Ph2SiF2 with KF and l8-Crown-6
(SD-2-5, 35, 49, 71, 89, 91, 121;
SD-3-3; NMR-1a)
A mixture of KF (0.29 g, 5.0 mmol) and 18-crown-6
(1.32 g, 5.00 mmol) in 8-10 ml of toluene was placed in a
plastic cap vial fitted with a magnetic stirring bar. Ph2SiF2
(1.20 g, 5.00 mmol) was added. The vial was capped and the
reaction mixture allowed to stir overnight, although
precipitation generally occurred within three hours. Gravity
filtration and washing with anhydrous ether produced a white
crystalline material (2.54 g, 94^ yield) that softens but does
not melt. After two recrystallizations from THF/ether,
analytically pure material results (mp 130-1 C). The 18-
crown-6 potassium salt of trifluorodiphenylsiliconate
[(K-18-C-6) (Ph2SiF3) ] was characterized as follows:


67
Elemental Analysis
Calculated for C2H3Si06F3K: C, 53.12; H, 6.31.
Found: C, 52.76; H, 6.35.
*H NMR 6 [multiplicity, area, assignment]
3.62 [s, 27, 0CH2];
7.00-7.22 [three-peak m, 6, meta and para Ar];
7.93~8.13 [four-peak m, 4, ortho Ar].
A broad water peak at 2.79 ppm was also present most of the
time.
13C NMR 6 [multiplicity, assignment]
70.77 [s, 0CH2];
126.95 [br s (two peaks?), ortho and para Ar];
127.77 [s, meta Ar];
138.58 [q ( 19F NMR 6 [multiplicity]
-109.86 [s (no JFSi evident, BWHM 53 Hz)].
29Si NMR 3 [multiplicity]
-92.00 [q (JS.F = 250 Hz)].
Some procedural changes are of interest, but are not
yet fully explored. When hexamethyldisilazane1*7 (1.60 g, 10.0
mmol) is added to the reaction medium and the white crystal-
line material is washed with cold CCl* instead of ether, then
a product (2.44 g, 89? yield) results which, without recrys-
tallization, has a melting point of 120-1 C. In addition,
hot EtOH may be used to recrystallize the product made in the


68
normal manner to yield material with a sharp melting point of
120 C. Finally, the reaction has been run at 50 C (oil
bath) and at 0 C (ice bath) and by using ether, instead of
toluene, as a solvent. A large amount of product (isolated by
filtration) was produced in all cases.
It is also worth noting that if the reaction is run
using a catalytic quantity of l8-crown-6 (SD-2-33) instead of
a stoichiometric amount, then, in contrast, only a small
amount of precipitate develops.
4.2.3 Low Temperature NMR Data for Ph2SiF3
(NMR-1)
The variable-temperature 19F NMR spectra for Ph2SiF3
were recorded as presented in Table IV. We have assumed a TBP
geometry for the anion in which the phenyl groups occupy two
of the equatorial positions.
In order to minimize viscosity broadening at low
temperature, the solvent for the spectrum at -77 C was an
acetone-d6/vinyl chloride mixture (ca. 50/50 by volume). All
other spectra were registered in pure acetone-d6.
Only successive changes are indicated in Table IV.
That we do not see F-Si coupling above -53-7 C is, we
believe, a result of the bandwidth of the signals and not
indicative of intermolecular exchange.
J


69
Table IV. Temperature-Dependent Behavior of Ph,SiF3
19F NMR 6
T (C) _______Apical F's_______ Equatorial F
-77.0 -100.54 doublet J(FSiF) = 2.6 Hz J(FSi) = 253 Hz BWHM = 6.0 Hz Integral: 2 -133.87 triplet J(FSiF) = 2.6 Hz J(FSi) = 205 Hz BWHM = 6.9 Hz Integral: 1
-68.6 -100.37 singlet BWHM = 6.9 Hz -134.03 singlet BWHM =8.6 Hz
-61.7 -100.22 BWHM = 8.6 Hz -134.13 BWHM = 14.6 Hz
-53.7 -100.01 BWHM = 14.8 Hz -134.7 no J(FSi) BWHM = 49 Hz
-43.6 -99.87 no J(FSi) BWHM = 76 Hz -134.37 BWHM = 142.7
-33.1 -99.76 BWHM = 244 Hz -134.62 BWHM =450 Hz
-22.0 -100.11 BWHM = 616 Hz -134.25 BWHM = 19700 Hz
-16.3 Signals are close to coalescence.
-12.0 -107.80 BWHM = 1180 Hz Coalescence temperature.
0.3 -110.44 BWHM = 520 Hz
13.5 -110.71 BWHM 266 Hz
37.7 -110.65 BWHM = 55 Hz


70
4.2.4 NMR Data for MePhSiF2
(NMR-2a) '
*H NMR 6 [multiplicity, area, assignment]
0.63 [t (JHCSiF = 6.10 Hz), 3, Me];
7.40-7.80 [m, 5, Ar].
19F NMR 6 [multiplicity]
-136.10 [q (JFSiCH = 6.02 Hz, Jps. = 290 Hz)].
4.2.5 Reaction of HePhSiF2 with KF and l8-Crown-6
(SD-2-11, 77, 79; NMR-2a)
A mixture of KF (0.29 g, 5.0 mmol) and 18-crown-6
(1.32 g, 5.00 mmol) in 10 ml of toluene was placed in a plas-
tic cap vial fitted with a magnetic stirring bar. MePhSiF230
(0.79 g, 5.0 mmol) was added. The vial was capped and the
reaction mixture was allowed to stir for four days, although
overnight is probably all that is necessary. Precipitation
generally occured within three hours. Gravity filtration and
washing with anhydrous ether produced a white crystalline
material (2.1 g, 85? yield) that softens but does not melt.
After recrystallization from hot THF, analytically pure
material results (mp 1331335 C). The l8-crown-6 potassium
salt of trifluoromethylphenylsiliconate
[(K18C6) (MePhSiF3) ] was characterized as follows:
Elemental Analysis
Calculated for Cx9H32Si06F3K: C, 47.48; H, 6.71.
Found: C, 47.45; H, 6.67.


71
*H NMR 6 [multiplicity, area, assignment]
0.07 [s, 3, Me];
3.62 [s, 29, OCHJ;
7.02-7.19 [three-peak m, 3, meta and para Ar];
7.80-8.00 [four-peak m, 2, ortho, Ar].
13C NMR 6 [multiplicity, assignment]
4.02 [very weak s (seen only once), Me (if signal is real)];
70.80 [s, 0CH2];
126.80 [broad s (two signals?), ortho and para Ar];
127.47 [s, meta Ar];
138.12 [s, ipso Ar].
13C NMR 6 at T = -77 C [multiplicity, assignment]
H.84 [t of d (JcslF(a) *9 Hz, JCSiF(e) 10 Hz), Me];
70.85 [s, 0CH2];
127.55 [s, ortho Ar];
128.70 [s, para Ar];
129.64 [s, meta Ar];
139.08 [t (JCSiF(a) = 7-3 Hz) iPS0 Ar^-
Small impurity peaks were evident at 129.64, 147.82 and
148.40 ppm. The last two appeared to be the peaks of a
doublet (A = 11.6 Hz). The sample for this spectrum was
an old solution.
19F NMR 6 [multiplicity]
-100.61 [s (BWHM = 130 Hz)].


72
29Si NMR 6
No resonance was observed at ambient probe temperature.
29Si NMR 6 at T = -77 C
We have not yet tried this.
Not surprisingly, we have also used CsF (0.76 g, 5.0
mmol) to make the trifluoromethylphenylsiliconate (2.3 g, 80$
yield). However, no product characterization has yet been
carried out.
4.2.6 Low Temperature NMR Data for MePhSiF3
(NMR-2)
The variable-temperature 19F and *H NMR spectra for
MePhSiF3 were recorded as presented in Table V. We have
assumed a TBP geometry for the anion in which the organic
ligands occupy two of the equatorial positions.
In order to minimize viscosity broadening at low
temperature, the solvent for the spectrum at -77 C was an
acetone-d6/vinyl chloride mixture (ca. 50/50 by volume). All
other spectra were registered in pure acetone-d6.
Only successive changes are indicated in Table V.
That we do not see F-Si coupling above -61.7 C is, initially
at least, a result of the bandwidth of the signals and not
indicative of intermolecular exchange.


73
Table V. Temperature-Dependent Behavior of MePhSIF,
19F NMR 5 XH NMR 6
T (C) Apical F's ~ Equatorial F_______________Methyl Region
-77.0 -83.63 quartet of doublets J(FSiF) = 2.6 Hz J(FSiCH) = 9 Hz J(FSi) = 251 Hz BWHM < 18 Hz Integral: 2 -132.93 singlet no J(FSiF) no J(FSiCH) J(FSi) =210 Hz BWHM < 9 Hz Integral: 1 0.01 triplet J(HCSiF ) = 9 Hz cL
-61.7 -83.53 singlet no J(FSiF) no J(FSiCH) BWHM = 29 Hz -133.32 singlet BWHM = 24 Hz ' 0.01 quartet J(HCSiF) = 6.6 Hz The central peak of the triplet split into two peaks.
-53.7 -83.47 no J(FSi) BWHM = 52 Hz -133.45 no- J(FSi) BWHM = 74 Hz No change.
-43.6 -83.49 BWHM = 117 Hz -133.58 BWHM = 167 Hz No change.
-33.1 -83.76 BWHM = 317 Hz -133.73 BWHM = 436 Hz Some broadening of the bands of the q.
-12.0 Signals are close to coalescence. Broader still.
1.8 -98.93 BWHM = 978 Hz Coalescence temperature. This is probably the upper limit. It is hard to establish exactly because of the huge 6 difference. Very broad singlet.
13.5 -100.3 BWHM = 529 Hz 0.05 singlet Broad, but starting to narrow some.
37.7 -100.61 BWHM = 130 Hz 0.07 Singlet, very sharp.


7^
Water was occasionally seen in the *H NMR spectra
throughout the entire temperature range. It's chemical shift
changed gradually with temperature from 2.75 ppm (37.7 C) to
3.85 ppm (-77.0 C) and seemed to intensify with decreasing
temperature. It is possible that water was condensing on the
outside of the NMR tube.
Several methods were attempted as means of removing
adventitious water or other impurities (like fluoride ion).
These included addition of HMDSlf7 and subjecting the sample
solutions to successive freeze-pump-thaw cycles. Nothing we
tried seemed to have a major effect on the observed rates of
exchange. However, occasionally, anomalous results were seen
(e.g., a huge, very broad 19F NMR Me3SiF peak at ambient
temperature after addition of HMDS). Though we tried very
hard, a more-thorough study of the possible effects of im-
purities should be attempted.
^.2.7 Simulation of the 19F_NMR Spectra of
Ph2SiF3~ and MePhSiF3
(NMR-1a, 2a)
We have simulated these spectra using a computer
program69 that plots calculated, time-dependent synthetic
spectra for a particular value of the mean lifetime t which is
defined,70 for a first order reaction, by (eq 27):
2k = 2/t = 1/t + 1 /tk =
a b
sec
i
(27)


75
Thus, theoretical curves were drawn for different
trial values of t and the best fit (by-visual superposition)
between simulated and experimental bandshapes then yielded the
appropriate t value at a particular temperature. Spin-spin
coupling was ignored since it is so small compared to the huge
differences in chemical shift. This analysis, combined with
the equations below, is known as the complete bandshape (CBS)
method. 70* 71 72
The spectra that were simulated were recorded in pure
acetone-d6. When vinyl chloride was added at intermediate
temperatures (e.g., -M3.6 C), the bandshapes typically became
narrower by ca. one-third. This could represent a slowing of
the exchange rate due to a decrease in solvent polarity (which
seems unlikely), or a removal of viscosity broadening, even at
these intermediate temperatures. In retrospect, if the latter
is true, then we have not properly considered viscosity
broadening in the bandshape analysis.
A plot of log k vs. 1/T provides the Arrhenius activa-
t
tion energy (E but commonly noted as AG ) and the pre-
exponential factor (A) since, experimentally, the increase in
rate constant (k) with temperature obeys the Arrhenius equa-
tion (eq 28):
log k = (-E /R-2.303) 1/T + log A
cl
= (-E /M.576 cal/K*mol) *1/T + log A
a
(28)


76
Alternatively, a plot of log (k/T) vs. 1/T provides
AH^ and AS^ according to the transition state theory of Erying
(eq 29):
k = ic(kbT/h) e(-AHt/RT) e(ASf/R) (29)
log k/T = (-AH+/R-2.303) 1 /T +
(ASt/R*2.303) + log (kb/h) + log k
= (-AHt/4.576 cal/K*mol) 1/T +
(ASt/4.576 cal/K*mol) + 10.318 + log k
where: R = 1.987 cal/K*mol
kb = 1.381*10 16 erg/K = Boltzman's constant
h = 6.626*10 27 erg*sec = Planck's constant
V
The value of the transmission factor (k) is dependent
on the multiplicity of the reaction path. When the reaction
path is simple, k = 1. For an n-path reaction, k = 1/n.
The x values and the appropriate plotting parameters
are given in Table VI for Ph2SiF3 and MePhSiFa Figures I
and II give the plots used to determine the activation
parameters which are listed in Table VII. We have assumed
that k = 1.


77
Table VI. CBS Analysis for Ph,SiF3 and MePhSlF,
T (C) -53.7 -43.6 -33.1 -22.0 -16.3 -12.0 0.3 13.5 37.7
T (K) 219.5 229.6 240.1 251.1 256.9 261.1 273.4 286.7 310.9
1/T MO3 4.556 4.355 4.165 3.982 3.893 3.830 3.658 3.488 3.216
Ph,SiF,"
T *105 800 250 73-0 27.0 16.0 11.0 3.50 1.50 0.300
k +103 0.125 0.400 1.37 3.70 6.25 9.09 28.6 66.7 333
log k 2.10 2.60 3.14 3.57 3.80 3.96 4.47 4.82 5.52
log k/T -.245 0.241 0.756' 1.17 1.39 1.54 2.02 2.37 3.03
MePhSiF,
T 1 0s 350 150 50.0 25.0 10.0 4.00 2.00 0.450
k +103 0.286 0.666 2.00 4.00 10.0 25.0 50.0 222
log k 2.46 2.82 3.30 3.60 4.00 .4.40 4.70 5.35
log k/T 0.115 0.463 0.921 1.20 1 .58 1.96 2.24 2.85
Table VII. Activation Parameters for PhgSiF3 and MePhSiF,
Ph,SlF, MePhSiF,
AG* (kcal/raol) 11.7 9.9
AH* (kcal/mol) 11.2 9.4
AS* (cal/Kmol) 2.5 -4.4
Av (Hz) 2508 3710
/\ 1 o 0) CO o 9090 25000
To (K) 261.1 275.0


log k
78
1/T E3
Figure I. Plot of log k vs. 1/T, using the data of Table VI,
for Ph2SiF3- (o, slope = -2561 K, intercept = 13-77,
correlation coefficient = -.9996) and MePhSiF3- (x, slope =
-2166 K, intercept = 12.29, correlation coefficient = -.9991).


79
1/T E3
Figure II. Plot of log k/T vs. 1/T, using the data of Table
VI for Ph2SiF3- (o, slope = -2451 K, intercept = 10.93.
correlation coefficient = -.9997) and MePhSiFa- (x, slope =
-2049 K, intercept = 9.42, correlation coefficient = -.9992).


80
Generally, it should be recognized that both statisti-
cal and systemic errors contribute to the overall error in AH*
and AS*. Good plot linearity is not proof of the reliability
of the results since weak statistical deviations may be accom-
panied by large systematic errors. Such errors may result
from bad temperature calibration or unrecognized temperature
dependency of site frequencies and/or bandwidths.
The accuracy of our temperature calibration is prob-
ably no better than 2 C. The chemical shifts contained in
Tables IV and V have been adjusted at every temperature using
an external CFCl3/acetone-d6: solution (ca. 50/50 by volume).
This solution is probably too concentrated for strictly proper
chemical shift calibration.
4.2.8 Preparation of PhSiF3
(SD-2-55, 59, 107; NMR-3)
To a 250 ml flask equipped with a magnetic stirring
bar and a condenser was added finely divided SbF3 (53.6 g, 300
mmol) under a blanket of Nz. This was cooled to -78 C with
dry ice/acetone and PhSiCl3 (63.5 g, 300 mmol) was rapidly
added by syringe. The solution froze fairly rapidly and was
allowed to come to ambient temperature overnight. The dark
reaction mixture was left stirring for an additional day and
then filtered and short-path vacuum distilled (bp 40 C/70
torr) to yield PhSiF3 (17.7 g, 36.0? yield), which was charac-
terized as follows:21


81
NMR 6 [multiplicity]
7.36-7.94 [m].
19F NMR <5 [multiplicity]
-141.06 [s (JFS. = 266 Hz].
4.2.9 Reaction of PhSiF3 with KF and l8-Crown-6
(SD-2-65; NMR-3)
A mixture of KF (0.073 g, 1.3 mmol) and l8-crown-6
(0.33 g, 1.3 mmol) in 2.5 ml of toluene was placed in a plas-
tic cap vial fitted with a magnetic stirring bar. PhSiF3
(0.20 gf 1.3 mmol) was added and the vial capped. The reac-
tion mixture was allowed to stir for three days, although
reaction is quite rapid. Filtration and washing with ether
produced a white crystalline material (0.56 g, 9055 yield) that
softens but does not melt. Attempts to recrystallize from THF
and THF/ether failed. The 18-crown-6 potassium salt of
tetrafluorophenylsiliconate [(K-18-C-6) (PhSiFa) ] was charac-
terized as follows:
Elemental Analysis
Calculated for Cx8H23Si06FlfK: C, 44.61; H, 6.03.
Found: C, 42.44; H, 5.64.
Dry loss: 1 .2456.
1H NMR 6 [multiplicity, area, assignment]
3.64 [s, 30, 0CH2];
7.05-7.25 [three-peak m, 3, meta and para Ar];
7.85-8.05 [four-peak m,' 2, ortho Ar].


82
19F NMR 6 [multiplicity]
-119.62 [s (no Jpgi evident, BWHM = 6.44 Hz)].
4.2.10 Low Temperature NMR Spectra of PhSiF*
(NMR-3)
19F NMR spectra were recorded from -12 to <-120 C. A
50% acetone-d6/vinyl chloride solution was used for the very
low temperatures. The fluorine resonance narrowed (and
silicon-fluorine satellites appeared) on going from ambient to
-12 C. Then, the spectrum remained constant throughout the
entire temperature region, except .for some broadening at the
low-temperature extreme.
19F NMR 5 at T = -44 C [multiplicity]
-119.31 Cs (JFgi = 206 Hz, BWHM = 1 Hz)].
4.2.11 NMR Data for Ph3SiF
(NMR-4)
The spectra for a commercial sample were recorded as:
*H NMR 6 [multiplicity, assignment]
7.30-7.80 [m, meta, para and ortho Ar].
13C NMR S [multiplicity, assignment]
129.50 [s, meta Ar];
132.00 [s, para Ar];
135.90 [s, ipso Ar];
136.00 [s, ortho Ar].
A slight impurity was evident at 134.00 ppm.


83
19F NMR 6 [multiplicity]
-169.90 [s (JFS. = 281 Hz)].
The I3C NMR assignments were made by analogy to pub-
lished data for Ph3SiH rather than to Ph3SiCl, for which the
ortho and ipso assignments are reversed.73
4.2.12 Reaction of Ph3SiF with KF and l8-Crown-6
(SD-2-25, 113, 115, 117, 125; NMR-4)
A mixture of KF (0.29 g, 5.0 mmol) and l8-crown-6
(1.32 g, 5.00 mmol) in 10 ml of toluene was placed in a plas-
tic cap vial fitted with a magnetic stirring bar. Ph3SiF
(1.39 g, 5.00 mmol) was added and the vial capped. If the
reaction mixture was allowed to stir for one day, then filtra-
tion and washing with cold ether produced a smaller-than-
typical amount of white crystalline material (1.0 g, 33%
yield). After six days, however, essentially quantitative
conversion occurred (2.7 g, 90% yield). The white crystalline
solid softens but does not melt and has not, to date, been
successfully recrystallized from THF, THF/ether, toluene,
methanol, ethanol or dichloromethane. The l8-crown-6 potassium
+ -
salt of difluorotriphenylsiliconate [(K-18-C-6) (Ph3SiF2) ]
was characterized as follows:
Elemental Analysis
Calculated for C30H39Si06F2K: C, 59.97; H, 6.54.
Found: C, 59.39; H, 6.39.


84
NMR 6 [multiplicity, area, assignment]
3.63 [a, 24.5, 0CH2];
7.00- 7.22 [four-peak m, 6, meta and para of two Ar rings];
7.30-7.75 [m, 5, meta, para and ortho of one Ar ring];
7.95-8.20 [four-peak m, 4, ortho of two Ar rings].
13C NMR 5 [multiplicity, assignment]
71.00 [s, 0CH2]5
128.00- 139.00 [10 peaks, Ar].
19F NMR 6 [multiplicity]
-99.50 [broad s (no JpSi evident BWHM ca. 200 Hz)].
A small impurity peak was seen once at -110.50 ppm.
Water was occasionally (but not always) observed in
the lH NMR spectrum. Attempts at simplifing the *H and 13C
NMR spectra by raising the probe temperature (to 50 C)
failed.
4.2.13 19F NMR Spectrum of a Ph3SiF2 /Ph3SiF mixture
(SD-2-57; NMR-4)
PhaSiF (0.03 g, 0.1 mmol) and Ph3SiF2 (0.06 g, 0.1
mmol) were combined in an NMR tube. A spectrum was recorded
overnight:
19F NMR 5 [multiplicity, area]
ca. -100 [very broad s (no dFSi evident), 3-6];
ca. -165 [very broad s (no JpSi
evident), 1].


85
4.2.14 Low Temperature NMR Data for Ph3SiF2
(NMR-4)
19F NMR spectra were recorded from 14 to -90 C. The
fluorine resonance narrowed (and silicon-fluorine satellites
appeared) on going from ambient to 0 C. Then, the spectrum
remained constant throughout the entire temperature region,
except for continued narrowing of the bandwidth from 0 to -50
C.
19F NMR 6 at T = -54 C [multiplicity]
-101.85 [s (JFS. = 255 Hz, BWHM = 3 Hz)].
4.2.15 NMR Data for (m-FPh)2PhSiF
(SD-2-61; NMR-6)
1H NMR 6 [multiplicity]
7.10-7^80 [m].
19F NMR 6 [multiplicity, area, assignment]
-169.40 [s (Jpsi = 282 Hz), 1, SiF];
-113.25 to -112.75 [m, 2, m-FAr].
A small impurity multiplet was seen from -114.25 to -113.89
ppm (integral = 0.1).
19F NMR <5 with lH decoupling [multiplicity, assignment]
-169.40 [s (JFS. = 282 Hz, no JFSiccCF)> SiF];
-113.-00 [s (no JFCCCSiF) m-FAr].
A small impurity singlet was observed at -114.00 ppm.


86
4.2.16 Low Temperature NMR Data for (m-FPh)2PhSiF
None were obtained for this compound.
4.2.17 Reaction of (m-FPh)2PhSiF with KF and l8-Crown-6
(SD-2-61; NMR-6)
A mixture of KF (.145 g, 2.50 mmol) and 18-crown-6
(0.66 g, 2.5 mmol) in 5 ml of toluene was placed in a plastic
cap vial fitted with a magnetic stirring bar. (m-FPh)2PhSiF7s
(0.79 gi 2.5 mmol) was added. The reaction mixture was capped
and allowed to stir for eighteen hours. Filtration and wash-
ing with ether produced a white crystalline material (1.6 g,
97? yield) that softens but does not melt. The l8-crown-6
potassium salt of bis(m-fluorophenyl)phenyldifluorosiliconate
C(K-18-C-6)+((m-FPh)2PhSiF2) ] was characterized as follows:
1H NMR 6 [multiplicity, area, assignment]
3.64 [s, 28, OCHJ;
6.65- 7.35 [m, 7, meta and para Ar];
7.65- 8.20 [m, 6, ortho Ar].
19F NMR 5 [multiplicity, area, assignment]
-119.00 to -117.80 [six-peak m, 2, m-FAr]5
-99.23 [s (JFSi = 258 Hz, BWHM = 14 Hz), 2, SiF].
Small impurity multiplets were observed at -114.30 and
-113.60 ppm, and a small, broad impurity singlet at -110.60
ppm.


87
19F NMR 5 with *H decoupling [multiplicity, assignment]
-118.3^ [s (no JFcccSiF^ m_FAr^5
-99.20 Is (JFS. = 258 Hz, BWHM est 10 Hz, no JFSiCCCp), SiF].
Small impurity singlets were seen at -114.30, -113*60 and
-110.60 ppm (broad).
4.2.18 Low Temperature NMR Data for (m-FPh)2PhSiF2
(NMR-6)
and 19F (coupled and decoupled) NMR spectra were
recorded from -20 to -95 C. The spectra were essentially no
different from those recorded at ambient temperature, except
that for this NMR solution the impurity peaks were not seen.
Additionally, it was later realized that the wrong
decoupling command (BB instead of HB) was used to record these
decoupled 19F spectra. The result was some strange splitting
patterns that we've chosen to ignore.
Subsequently, the following spectrum was recorded in
an 80? vinyl chloride/20? acetone-d6 solution at -110 C. We
didn't understand it at the time and have not pursued it
further.
19F NMR 6 with *H decoupling [multiplicity, area, assignment]
-143.50 [s, 0.4, not assigned];.
-116.00 [s, 1.7, not assigned];
-112.00 [s, 0.1, not assigned];
-111.00 [s, 1.0, not assigned];
-100.00 [s, 2.0, not assigned].